7KH7KUHDWIURPWKH3RSXOLVW/HIW
Kurt Weyland
Journal of Democracy, Volume 24, Number 3, July 2013, pp. 18-32 (Article)
3XEOLVKHGE\7KH-RKQV+RSNLQV8QLYHUVLW\3UHVV
DOI: 10.1353/jod.2013.0045
For additional information about this article
http://muse.jhu.edu/journals/jod/summary/v024/24.3.weyland.html
Access provided by KOC University (26 Jan 2015 14:10 GMT)
Latin America’s Authoritarian Drift
The Threat from the
Populist Left
Kurt Weyland
Kurt Weyland is Lozano Long Professor of Latin American Politics
at the University of Texas–Austin. He has published three books and
many articles and chapters on democratization, neoliberalism, populism, social policy, and diffusion in Latin America. His current research analyzes the wavelike diffusion of political-regime contention
across European and Latin American countries.
Around the turn of the millennium, prominent Latin America special-
ist Scott Mainwaring highlighted the surprising endurance of democracy
in that region after the transition wave of the late 1970s and 1980s.1 During that interval, no democracy had permanently succumbed to a military coup or slid back into authoritarian rule. After decades marked by
instability in numerous countries, especially Argentina, Bolivia, and Ecuador, this newfound democratic resilience came as a welcome surprise.
But at about the time Mainwaring was writing, onetime coupmaker
Hugo Chávez was winning election to the Venezuelan presidency and
beginning to move his country away from democratic rule. Venezuela
had survived the rash of military coups that swept the region in the
1960s and 1970s to become a byword for democratic stability in Latin
America. Economic deterioration, political ossification, and rampant
corruption had brought sustained decay, however, and paved the way for
this radical populist, former army officer, and would-be golpista (he had
led a violent putsch that failed in February 1992) to decisively win the
free and fair December 1998 balloting. Using plebiscitarian strategies
to transform the country’s liberal institutional framework, concentrate
power, and entrench himself, Chávez set about strangling democracy
and putting competitive authoritarianism in its place.2 He remained as
president till he died of cancer on 5 March 2013.
The Chávez phenomenon has had strong demonstration and contagion effects beyond Venezuela. Eager to overcome instability and cement their own supremacy, Presidents Evo Morales of Bolivia (2006–)
Journal of Democracy Volume 24, Number 3 July 2013
© 2013 National Endowment for Democracy and The Johns Hopkins University Press
Kurt Weyland
19
and Rafael Correa of Ecuador (2007–) have emulated Chávez’s script.
As did their political ally and financial benefactor, they have used constituent assemblies to augment executive powers, allow for presidential
reelection, and weaken institutional checks and balances. From that position of strength, they have made discretionary use of the law for political purposes. With this discriminatory legalism, they have attacked, undermined, and intimidated the opposition in their respective countries,
moving toward competititve authoritarianism as well.
Similarly, strong informal pressures and disrespect for constitutional
principles have enabled Daniel Ortega (2007–) to establish his hegemony in Nicaragua. President Manuel Zelaya of Honduras (2006–2009)
also sought to follow in the footsteps of Chávez, Morales, and Correa by
convoking a constituent assembly and preparing his own perpetuation
in power; yet coordinated opposition from Congress, the courts, and the
military aborted this effort through a controversial June 2009 coup. Even
President Cristina Fernández de Kirchner of Argentina (2007–), whose
fervent supporters take inspiration from Chávez, is eyeing constitutional
changes and renewed reelection (she is now in her second term). Given
Argentina’s weak and disunited opposition, this push for entrenchment,
combined with continuing attacks on the press and the president’s personalistic command over the state, has created alarm in civil society
about looming threats to the country’s hard-won democracy.
That Venezuela had already fallen under nondemocratic rule was
confirmed in October 2012 by Chávez’s unfair reelection, achieved with
the help of intimidation tactics, tight restrictions on the opposition, and
the massive misuse of the state apparatus. Since the third wave reached
Latin America in 1978, the region had seen only occasional threats and
temporary interruptions of democracy in individual nations. The recent
suffocation of political pluralism in a whole group of countries is without precedent. For the first time in decades, democracy in Latin America
is facing a sustained, coordinated threat. The regional trend toward democracy, which had prevailed since the late 1970s, has suffered a partial
reversal. Unexpectedly, democracy is now on the defensive in parts of
the region.3
With its electoral façade and progressive rhetoric about helping the
excluded, the soft authoritarianism that is taking hold in parts of Latin
America has an attractive face. It exerts an appeal on regional and global
public opinion to which academics are not immune. The military dictators of the 1960s and 1970s were ogres with no legitimacy who depicted
themselves as stopgaps—house cleaners putting politics in order so democracy could return. By contrast, Chávez and friends have claimed
to institute a new participatory—and hence qualitatively better—form
of democracy and to promote social equity and national independence.
Rather than a short-lived detour, they seek to carve out a distinct development path purportedly leading to what Chávez called “socialism for
20
Journal of Democracy
the twenty-first century.” Their competitive authoritarianism appears
not as a limited interruption but a permanent alternative to pluralist,
representative democracy. This appeal is unusual among contemporary
nondemocracies; it contrasts with Russian strongman Vladimir Putin’s
more bluntly unsavory brand of autocracy, for instance. These “progressive” claims aggravate the risks emanating from the recent turn to
authoritarian rule.
The current authoritarian trend in Latin America is not regionwide:
Major countries such as Brazil, Chile, Mexico, and now Colombia seem
safely consolidated as democracies; Costa Rica and Uruguay boast especially high democratic performance. But the unexpected ease with which
a coordinated nucleus of competitive authoritarianism has emerged must
give pause. To see even Argentina, with its tragic history, being lured
by the siren song of personalistic plebiscitarianism is worrisome indeed.
Establishing Political Hegemony
As Steven Levitsky and James Loxton and Raúl Madrid have emphasized, Chávez and his friends used populism to entrench their predominance and install competitive authoritarian regimes.4 Populism,
understood as a strategy for winning and exerting state power,5 inherently stands in tension with democracy and the value that it places upon
pluralism, open debate, and fair competition. Populism revolves around
personalistic leadership that feeds on quasi-direct links to a loosely organized mass of heterogenous followers. Bypassing or subjugating intermediate institutions such as firmly organized parties, the leader—
often a charismatic figure—establishes face-to-face contact with large
numbers of citizens. In earlier decades, mass rallies were crucial; nowadays, television allows populists to reach their followers “in person.”
Chávez hosted a regular Sunday talk show. The leader in turn ascertains
“the people’s will” through frequent popular votes and opinion polls.
To show vigorous leadership, seem indispensable, and boost followers’
loyalty, populist politicians are fond of constantly attacking enemies, at
least rhetorically. In this way, the leader blames others for the problems
that have allowed the leader to take power and act as the savior of the fatherland. The leader is the star of a drama in which “the people” struggle
heroically under the leader’s direction against selfish, corrupt enemies
at home and abroad.
As a political strategy, populism can have variegated and shifting
ideological orientations and pursue diverse economic and social policies. Contemporary Latin America has seen populist presidents from
the right, such as Argentina’s Carlos Saúl Menem (1989–99) and Peru’s
Alberto Fujimori (1990–2000), and populists of the left such as Chávez,
Morales, and Correa. Many populist leaders have embraced economic
nationalism and state interventionism, yet others have imposed free-
Kurt Weyland
21
market reforms. In a particular twist, the Peronist Menem dismantled the
protectionism-based developmental model that his own party’s populist
founder, Juan Perón (president from 1946 to 1955, and then again from
1973 to 1974), had installed.
Populism will always stand in tension with democracy. The logic of
personalism drives populist politicians to widen their powers and discretion. Because these leaders sustain their influence via personal appeals
rather than intermediary organizations, they see any institutions outside
their control as obstacles to be bypassed or overcome. Determined and
politically compelled to boost their personal predominance, populist
leaders strive to weaken constitutional checks and balances and to subordinate independent agencies to their will. They undermine institutional protections against the abuse of power and seek political hegemony.
Correspondingly, populist leaders treat opponents not as adversaries
in a fair and equal competition, but as profound threats. Branding rivals
“enemies of the people,” they seek all means to defeat and marginalize
them. Turning politics into a struggle of “us against them,” populists undermine pluralism and bend or trample institutional safeguards. Populist
leaders also put strong pressure on independent forces in civil society
and strive to control the media, especially television. All these attacks,
depicted as a defense of the people against rapacious elites, are also
meant to strengthen leader-follower bonds and thus to compensate for
the lack of organizational mediation. The absence of institutional discipline in the populist movement prompts the leader to recharge the base’s
loyalty through heroic activism. In all these ways, the populist notion of
politics as an “all or nothing” struggle damages democracy.
Populism, whether of the left or the right, is a threat to democracy.
Yet in Latin America today, the graver and more sustained danger is
coming from the leftist variant. Chávez set the model. As soon as he was
elected president of Venezuela, he set about revamping the country’s
institutional framework. First, he called a constituent assembly. Then,
to dislodge the established political class that he charged with selfishness and corruption, he successfully pushed to close the recently elected
bicameral Congress, where his followers held only about a third of the
seats. Thanks to a reengineered electoral system, Chávez dominated the
constituent assembly that boosted his powers, ended the ban on consecutive terms, and created a new unicameral (and hence easier to control)
national legislature. These institutional victories—plus the promise of
socioeconomic change—lifted Chávez and his camp to victory in the
2000 elections. Moreover, he took control of the courts and other independent institutions, such as Venezuela’s electoral commission, and
soon had a stranglehold on all branches of government.6
Chávez and his supporters, along with some academics and intellectuals, claimed that Venezuela had become a participatory democracy.
Common citizens, so long neglected by traditional politicians, could
22
Journal of Democracy
at last have a direct say in their own governance. There is some truth
to these claims when it comes to local decision making and socialprogram implementation, but they are unconvincing as applied to the
crucial arena of national policy making.
There should be no mistaking that Hugo Chávez made every important decision and thoroughly determined his country’s political
course. No aide could rein him in, and the people lacked the capacity
to advance their collective will independently. The absence of firm
popular organization and of transparent decision-making procedures
precluded effective bottom-up influence. Political initiative emanated
from the leader, not the citizens. Chávez never changed any significant plan due to popular resistance. Even when he lost, as in the 2007
constitutional plebiscite, he simply redoubled his efforts and pushed
through to his goals. Rather than driving decisions, the populace was
the object of Chávez’s populist strategies and tactics, as can be seen
from the rapid rise and decline of chavista movements such as the
Bolivarian Circles. Talk about direct democracy cannot change contemporary Venezuela’s status as a prototypical case of personalistic
populism. Chávez’s handpicked successor Nicolás Maduro, who won
an April 2013 special election to the presidency, is perpetuating this
top-down style—witness the strikingly opaque machinations that surrounded Maduro’s assumption of presidential powers during the later
stages of Chávez’s illness.
Chávez’s success in revamping Venezuelan politics and fortifying
his personal dominance turned his strategy of constitutional reform into
a script that other populist-leaning left-wing leaders followed. The core
of the Chávez method is to use plebiscitarian mass support in order to
transform established institutions, dismantle checks and balances, concentrate power in the hands of the president, and promote immediate
reelection. Like their Venezuelan role model and generous patron, Evo
Morales in Bolivia,7 Rafael Correa in Ecuador, and Manuel Zelaya in
Honduras (till he was stopped) called for constituent assemblies with
the aim of boosting the presidency’s powers and paving the way toward
indefinite reelection to that office. Bolivia and Ecuador’s respective
histories of acute instability—including interrupted presidencies—and
consequent hopes for “a fresh start” guaranteed strong popular support
for the new chief executives. With this majoritarian backing, personalistic leaders undermined liberal, pluralist institutions.
In Bolivia, the Morales government shut the opposition out of decisive stages of the constitution-drafting process.8 The charismatic leader
then won his foes’ agreement to a referendum on the tailor-made charter by promising not to run in 2014. But he soon went back on this
vow; a typical populist, he is determined to cling to power. In Ecuador,
Rafael Correa got his constituent-assembly election by engineering the
irregular removal of more than half the members of Congress. By in-
Kurt Weyland
23
voking popular sovereignty, this populist leader managed to defeat his
adversaries and rewrite the rules via a new charter that greatly increased
presidential powers.
The Weapon of Discriminatory Legalism
Once these populists of the left established predominance, they used
their unfettered control over all branches of government to limit debate,
strike at opponents, and drastically tilt the electoral playing field. These
maneuvers dismantled democratic accountability and eliminated safeguards against arbitrariness. Hegemonic presidents called frequent referenda to garner plebiscitarian acclaim, but always with arrangements in
place to ensure that these ballot-box exercises never gave the opposition
a fair chance to win. When adversaries did manage to claim a victory,
as happened occasionally from 2007 to 2010 in Venezuela, Chávez employed all kinds of shenanigans to render it meaningless. In late 2010, for
instance, he crippled a newly elected parliament with significant opposition representation by having the outgoing assembly, where his supporters
had exclusive control, delegate extensive legislative powers to him.
In these ways, left-wing populists have slowly but surely smothered
democracy and entrenched competitive authoritarian rule in several Latin
American states. Their brand of soft authoritarianism violates basic principles of democracy by placing controls on the media and the opposition while the government electioneers using state resources. Even when
presidents command high popularity, as left-wing populists often have,
contests held under such profoundly unfair conditions cannot qualify as
democratic. Where the parameters of political choice are so badly distorted, majority support cannot compensate for serious infringements of
pluralism and competitiveness.
While justifying their undemocratic moves with progressive claims,
left populists have eagerly availed themselves of timeworn tactics of
Latin American politics. Presidents in the region have long been known
for efforts to distort electoral competition and unfairly perpetuate themselves in power. In particular, they have applied discriminatory legalism and its maxim “For my friends, everything; for my enemies, the
law!” As populist chief executives have commandeered all major institutions including the courts, they have used formally legal authority
in discretionary ways to promote their cronies and allies while punishing or intimidating critics and opponents in politics and society. With
the government controlling all avenues of appeal and avoiding blatant
violations of formal rules, those targeted find few chances for domestic
recourse or the gathering of international support.
Here again, Chávez proved himself a trendsetter: He showed how skillfully an elected incumbent can employ discriminatory legalism to stifle
debate and push critics and opponents to the wall. With comprehensive
24
Journal of Democracy
control over Venezuela’s political institutions, Chávez closed a number
of independent television stations and threatened the remaining ones;
used trumped-up charges to jail or drive into exile recalcitrant judges
and opposition leaders; and exploited oil rents and the state apparatus for
campaigning. In these ways, he sapped the opposition’s chances of success and ensured himself frequent victories at the polls. If his adversaries
did win against all odds, he used various ploys to limit the effects. After
the opposition managed to win the mayoralty of Caracas in 2008, for instance, Chávez folded much of the city into a new Capital District under
a handpicked commissioner who was given most of the power and funding that had previously been under the mayor’s control. With such unfair
tactics, this populist leader undermined democracy and skewed political
competition.
Seeing how discriminatory legalism has served to entrench competitive authoritarian rule in Venezuela, the leftist presidents of Bolivia, Ecuador, and Nicaragua have followed suit and imitated Chávez. In Latin
America today, the strangling of pluralism and competitiveness is not
confined to a single case. Instead, formally legal means to control the
media, attack the opposition, and massively use the state for electioneering are catching on in a whole set of countries as handy expedients for
incumbents intent upon securing a lock on power.
In Bolivia, Evo Morales and his Movement Toward Socialism have
used trumped-up charges of administrative irregularities, corruption, terrorism, and genocide against numerous opposition politicians, imprisoning some, driving many others out of the country, and intimidating the
rest. The competitiveness that is essential to democracy cannot survive
in such a hostile setting. Ecuador’s Rafael Correa has applied similar
tactics, for example against the politician who challenged him in the
2006 election. Correa also seized on a 2010 police rebellion—painted by
him as a coup attempt—as a pretext for cracking down on independent
social and political forces. And he has intimidated the media by suing
for exorbitant damages and stiff prison sentences over an opinion piece.
Daniel Ortega has decreed many paralegal measures in Nicaragua’s
weakly institutionalized polity and has put persistent pressure on independent NGOs. After extracting concessions from an opposition leader
who had been convicted of corruption charges, Ortega packed the courts
and then had his appointees on the bench exempt him from the constitution’s ban on immediate reelection. Furthermore, Ortega’s supporters
relied on manipulation and fraud in the 2008 municipal elections. In Nicaragua, discriminatory legalism has shaded into systematic illegalism.
Even in Argentina, where democracy has so far survived populist
pressures, President Cristina Fernández de Kirchner (often known as
CFK) has started to follow a Chávez-like script. Businesspeople who
publicly criticize her have found themselves targets of special tax audits. Media outlets that draw her ire—the newspaper Clarín is a particu-
Kurt Weyland
25
lar thorn in her side—have faced everything from antitrust investigations to mob violence. Even as it has been bullying critics, the ruling
group around Kirchner has been floating the idea of calling a constituent
assembly to pave the way for a third CFK term. Argentine civil society,
however, has pushed back harder against this scheme than civil society
in a “Bolivarian” country would likely be able to do. Mass protests in
late 2012 noisily opposed the extension of CFK’s rule, suggesting that
Argentina will not easily be led down the Chávez path.
Populism Left and Right
The populist wing of Latin America’s contemporary left poses a significantly stronger challenge to democracy than did the wave of rightwing populist presidents who rose to prominence in the 1990s (or in Colombia’s case, the 2000s). Carlos Menem and Alberto Fujimori, along
with Brazil’s Fernando Collor de Mello (1990–92) and Colombia’s Alvaro Uribe (2002–10), also employed populist strategies, but on behalf
of neoliberal economic policies and, in Peru and Colombia, the need to
defeat violent leftist guerrillas. Despite differing from the current crop
of left-wing populists on ideology and policy, these rightist presidents
nonetheless favored a similar personalistic leadership style and mobilized amorphous, heterogeneous mass followings in a quasi-direct fashion. Each president cast himself as the people’s champion in a struggle
against malign forces such as established politicians and left-wing insurgents. In these ways, neoliberal populists garnered wide popular support
that they sought to sustain with plebiscitarian tactics.
In typical populist fashion, these neoliberal politicians sought to
boost presidential powers, weaken checks and balances, and extend their
control over the government while preparing their own reelections. Menem, for instance, bent constitutional rules by issuing an unprecedented number of “emergency” decrees and packed Argentina’s Supreme
Court in order to protect his arrogations of power. Collor steamrolled
Brazil’s Congress, forcing legislators to accept drastic macroeconomicstabilization measures by using his decree powers to confront the lawmakers with a fait accompli. Menem and Uribe pushed constitutional
changes designed to help them get reelected. Most blatantly, Fujimori
closed Congress and took control of the courts with his 1992 autogolpe
(self-coup). Faced with strong international protests, he sought to tack
away from naked authoritarianism by calling a constituent assembly that
augmented presidential prerogatives and allowed for his reelection. The
new charter also weakened the legislative branch by replacing Peru’s
bicameral Congress with a unicameral assembly. In these ways, rightwing populists damaged Latin American democracy, destroying it altogether for a time in Peru.
But this deterioration was limited in severity in Argentina, Brazil, and
26
Journal of Democracy
Colombia, and in duration in Peru. Collor did not keep his promise to
“kill inflation” and was defeated by Brazil’s political class, which forced
him to resign amid a corruption scandal. Menem and Uribe did achieve
policy success and parlayed the resulting popularity into convincing reelection victories. But the desire of each to win a third consecutive term
ran afoul of intraparty opposition in Argentina and a powerful independent Constitutional Court in Colombia. When Menem and Uribe stepped
down, democracy in Argentina and Colombia recovered. Even Fujimori,
who in 2000 managed to win a second reelection, fell soon thereafter as
his ever more extreme personalism collapsed under its self-destructive
logic. Once the president had pulverized the party system and subjugated
the Peruvian state, his rule was thoroughly extra-institutional, resting on
shady personal connections sustained by widespread corruption. When
evidence of this crass bribery surfaced, Fujimori’s hold on power vanished. The political demise of Fujimori—who is now serving jail time for
corruption and human-rights abuses—brought back full democracy, with
ample public debate and free and fair elections. Thus, right-wing populism did not ruin democracy in Argentina, Brazil, or Colombia, and in
Peru democracy’s destruction and temporary replacement were followed
by a quick resurrection.
By contrast, left-wing populism has a more negative balance sheet.
Chávez dominated Venezuelan politics for fourteen years, stopped only
by his death. His underlings have good chances of retaining control, aided
by the emotional impact of Chávez’s “martyrdom.” Morales, Correa, and
Ortega have also cemented their respective hegemonies and prepared their
own continuations in power. Bolivia’s president, as mentioned, has gone
back on his promise not to run again in 2014. With Correa’s February
2013 reelection to a third term now behind him, he is poised to tighten his
own political stranglehold. Given these leaders’ unfettered control over
state resources and their willingness to employ discriminatory legalism,
opposition forces face steep uphill battles in a context of heavily rigged
electoral competition. Incumbent governments have jailed opposition
politicians or driven them out of the country in Bolivia, and have attacked
and intimidated civil society in Ecuador and Nicaragua. These tightening
constraints on political pluralism give the nondemocratic leaders of leftwing populism ever firmer foundations for their rule.
The Levers of the Left
Why has left-wing populism been doing more damage to democracy
in Latin America than right-wing populism did? This asymmetry reflects differences not in intention, but in capacity. Today’s populists of
the left command greater political strength and have more policy tools.
They can push further down the road toward concentrated power than
could their neoliberal cousins of a few years ago.
Kurt Weyland
27
First, right-wing populism has a temporary (usually crisis-driven) support base, while leftist populism has more lasting roots, particularly in
the “informal” sectors that figure so largely in the economies of many
Latin American countries. Second, by reducing the power of the state over
markets and private economic actors, neoliberalism diminishes the power
of right-wing leaders. The growing state interventionism favored by leftwing populists, by contrast, gives them additional means of influence.
Third, neoliberalism exposes right-wing populists to international pressures for democracy; economic nationalism, by contrast, insulates leftist presidents from such exhortations. Finally, right-wing populists acted
separately, while today’s left-wing leaders form a coordinated group. This
cohesion further disarms international pressures to maintain democracy.
For all these reasons, Bolivarian leaders have managed to strangle democracy much more effectively than neoliberal populists ever could.
The populists of the right always stood on shakier political ground
than that of Chávez and his friends. Neoliberal populists won office
by vowing to solve crises. Success made these leaders dispensable. By
contrast, left-wing populists invoke structural problems—poverty, inequality, marginalization—that allow only for slow progress and resist
definitive resolution. Stubborn problems thus justify one reelection of
“the leader” after another. Moreover, these presidents have relied not
only on performance-based legitimacy, but also on durable identitybased appeals that cast them as champions of, for example, informal
workers, barrio residents, or indigenous people.
The right-wing populist presidents Menem, Fujimori, and Collor rose
to power amid bouts of hyperinflation. These economic catastrophes discredited the existing parties in Argentina, Peru, and Brazil, respectively,
opening space in each country for an outsider who pledged to stop the
pain. But the political weakness that followed Collor’s failure to end inflation contributed to his downfall on corruption charges. Menem and Fujimori eventually brought skyrocketing prices under control and received
massive popular support in return. But the backing did not endure: Once
these presidents had restored economic stability, voters switched to worrying about poverty and unemployment—problems that executives committed to neoliberal austerity, budget discipline, and privatization found
much harder to solve. Within Menem’s own Peronist party, for instance, a
rival running to Menem’s left cut him off from his hopes for a third term.
Fujimori and Uribe also won popular support with their success in fighting guerrillas. An improving security situation boosted each president’s
popularity for a while. But as the danger receded, especially in Peru, citizens’ priorities shifted, exposing the two chief executives to a paradox of
success. Their very accomplishments hamstrung their efforts to perpetuate
themselves in office. Fujimori fell in 2000, the victim of his achievements
as well as his considerable excesses, and Uribe failed to parlay his 2008
victories over leftist insurgents into another reelection in 2010.
28
Journal of Democracy
Left-wing populists, by contrast, base their appeal on structural problems. They highlight Latin America’s longstanding social deficits, especially widespread poverty and inequality. While the established political
class looks self-serving and beholden
to privileged elites, left-wing populists
project concern for common citizens
The neoliberal economics
to which recent right-wing and start generous social programs
that—despite frequent administrative
populists were devoted
problems stemming from politicizaended up diminishing their
tion—significantly increase benefits,
control over economic
alleviate destitution, and bring symmatters and hence weakbolic recognition as well. This deliberening them politically.
ate identification with ordinary people
and their plight is reinforced by the
leaders’ affiliations with the popular
sectors from which they spring (or with which they identify themselves).
Left-populist identity politics is especially important in Bolivia, where the
supporters of Morales like to boast that he is the first indigenous president that this majority-indigenous country has ever had. Similarly, Chávez
dwelt often on his humble upbringing and spoke in a popular (and vulgar)
idiom not previously associated with presidents of Venezuela.
Left-wing populists claim to be the first chief executives to embrace
a preferential option for the poor. Their social programs embody this
commitment, but cannot quickly overcome longstanding structural deficits. This slow progress with no end in sight yields more durable political payoffs than neoliberal populists’ success in solving dramatic crises.
Left-wing populists prove their social orientation and performance, and
then point to the difficulty of the task in order to explain why they must
stay in office. Thus, activist social policies further cement identity-based
loyalties. These bonds give left-wing populism more reliable political
sustenance than neoliberal leaders can command and allow left-wing
populists to do graver damage to democracy.
The neoliberal economics to which recent right-wing populists were devoted, far from fortifying their political hegemony, ended up diminishing
their control over economic matters and hence weakening them politically.
Certainly, in the short run market-based reforms can augment presidential
influence. Privatization programs, in which the government decides who
may buy public enterprises, offer obvious opportunities for extracting favors. But once firms pass into private hands, the government loses control.
Thus neoliberalism’s end product is reduced presidential clout.
Neoliberal orthodoxy limits leaders in other ways. Budget discipline restrains patronage spending. Personnel cuts shrink the leeway
for hiring cronies. Reliance on market forces precludes large-scale
employment programs. Moreover, business and international financial
institutions insist on firm, transparent legal parameters and thus reduce
Kurt Weyland
29
leaders’ autonomy and discretion. In sum, neoliberalism constrains
populist chief executives and hinders their continued reelections.
By contrast, left-leaning populists boost state interventionism. They
add to the public payroll, increase regulation, and nationalize enterprises.
This yields growing patronage resources, so presidents can buy support
and press their opponents. As ever more people come to depend on the
state, they become possible targets for discriminatory legalism. Citizens
have an incentive to toe the line and back the incumbent, however grudgingly, as in the 2012 election that returned a dying Hugo Chávez to the
presidency of Venezuela. Businesspeople need to think twice before funding oppositionists lest the government find a pretext to revoke business licenses, deny access to foreign exchange, or impose other sanctions. Once
a populist president has established hegemony and defanged accountability mechanisms, extensive state interventionism offers untold new chances
to reward friends, punish foes, and tilt the playing field.
In the years since the Cold War’s end, international pressures in favor
of democracy have come to the fore. Neoliberal economic-policy commitments exposed right-wing populists to these in ways that left-wing
populists have seldom if ever experienced. After Fujimori’s self-coup,
he quickly backed away from open authoritarianism lest economic sanctions foil his market reforms. To preserve his hard-won economic success, Fujimori called elections for a constituent assembly and restored
room for political competition. Neoliberalism trumped authoritarianism.
Similar external pressures later limited Fujimori’s efforts to manipulate
the 2000 presidential election; they also hindered Menem’s and Uribe’s
attempts to stay in office.
Left-wing populists, by contrast, can huddle behind economic nationalism. Reduced reliance on global market forces and rising statism build
walls against international efforts to promote democracy. Under fire
for blatant uses of discriminatory legalism, Chávez pulled out of hemispheric institutions such as the Inter-American Commission on Human
Rights. He also kept international election observers out of Venezuela,
which helped him to hide how badly he had warped the competitive
arena in his own favor. With the continuing boom in oil and naturalgas prices, commodity-rich Bolivia, Ecuador, and Venezuela have been
able to ignore global market pressures (as has Nicaragua, which receives
Venezuelan subsidies).
Yesterday’s right-wing populists differed from today’s left-wing populists, finally, in being less organized as a group. Neoliberal presidents may
have banded together to found the Southern Common Market (Mercosur)
but they never did much to support one another diplomatically. For instance,
neither Menem nor Collor backed Fujimori after his self-coup, and neither
ever came close to trying to shutter Congress. When Chávez put democracy
to death by constituent assemblies, he inspired imitators. Fujimori’s more
direct attack on democracy had no such effect on his neoliberal peers.
30
Journal of Democracy
Left-wing populists act in coordinated ways. Morales, Correa, and
Zelaya (who was stopped early in the process) sought to retrace Chávez’s
path through constitutional change to political hegemony and discriminatory legalism. Daniel Ortega took advantage of Nicaragua’s low level
of institutionalization to push his changes through by informal means.
They all benefited from Chávez’s petrodollars, political advice, diplomatic support, and security protection. This comprehensive backing
from Caracas strengthened left-wing populists both at home and abroad.
Thus did Chávez help to smother democracy in several countries.
The tendency of left-wing populists to close ranks also serves to protect their assaults on political competition from international rescue efforts. The hemispheric community can force the president of Peru to
retreat from open authoritarianism, but has no such leverage on a cohesive group of countries that aid one another and wield something akin
to a veto within regional institutions. Among their tacit allies have been
more moderate countries, such as Brazil, which see Bolivarian radicalism as a handy foil that raises their bargaining power vis-`a-vis Washington. The diplomatic self-interests of Latin American democracies have
thus played a role in hampering international efforts to prevent authoritarian backsliding in Venezuela, Bolivia, Ecuador, and Nicaragua.
In fact, left-wing populists have skillfully used the region’s democracy-defense arrangements to abet their own internal assaults on democracy. International mechanisms to protect competitive rule were designed
with dramatic threats, such as coups against elected presidents, in mind.
When Chávez faced an irresolute attempted coup in 2002, these mechanisms helped him, just as they helped Evo Morales when he had to deal
with mass protests in 2008. It is no small matter that chief executives,
who naturally display solidarity with their counterparts elsewhere, are
typically the ones who must apply these measures.9
Discriminatory legalism has so far proved a democracy-strangling
tactic that the international community has found hard to rein in. Outsiders to a country must first pierce the veil of formal legality, and then decide when discrimination has become bad enough and broad enough to
count as a violation of democracy. Left-wing populists typically move
gradually to undermine democracy; where is the threshold that calls for
international intervention? The most visible victims are usually legislators, high-court judges, and party politicians—not types that foreign
presidents will feel most eager to rescue. As elected populist presidents
squeeze and manipulate their opponents, diplomatic backing against the
onslaught can prove scarce.
Because they can so easily be made to shield perpetrators more than
victims, current democracy-protection protocols in the region are serving to undermine democracy and—however unintentionally—to further
tilt the playing field in several countries. Like discriminatory legalism
at home, the asymmetrical internationalism that informs regional coun-
Kurt Weyland
31
cils helps to spread and entrench nondemocracy. The new competitive
authoritarian regimes of Latin American leftist populism lack the harshness of old-school dictatorships, but they have achieved a degree of
“perfection” (to borrow Mario Vargas Llosa’s ironic term) that even
Mexico’s long-ruling PRI in its heyday could not rival.
A Surprising Threat
Historically, it has been the right that has done the most damage to
competitive civilian rule in Latin America, so when a new threat from
the left emerged during a time of what appeared to be democratic consolidation, many observers were surprised. For decades, oligarchs had
stifled mass participation while soldiers mouthing anticommunist slogans had all too often intervened to crush popular empowerment and
democracy. Leftists bore the brunt of the repression, learning to stop
calling democracy a “bourgeois farce” and to embrace human-rights
safeguards and checks on state power. Much of Latin America’s left has
thus come to have strong democratic credentials.
Populist politicians, however, lack firm commitment to ideologies and
principles and concentrate on the quest for personal power. The urge to
boost the leader’s clout, the dislike of constitutional limits, and the harsh
treatment of rivals make populism an inherent threat to democracy. Populists both right and left have displayed these tendencies, but the latter have
done more damage to democracy with their greater staying power and
more skillful efforts to hoard power, knock down institutional safeguards,
squeeze opponents, and skew competition. Beneath a veneer of formal
legality, these populists have blunted and even exploited the hemisphere’s
methods for guarding against reversals of democracy.
With its claims to make democracy more direct and to be especially
mindful of the poor, left-wing populism has crafted an attractive message.
It has spread from Venezuela to several other countries and has stimulated
interest elsewhere, especially Argentina. The temptations that it spawns
make Chávez-style populism a particular threat to democracy.
This threat also seems to have clear limits, however. Brazil, Chile,
Costa Rica, Mexico, Uruguay, and now Colombia boast stable democracies. Steady institutions, pluralist party systems, and respectable government performance leave less room for populists. The downsides of
Bolivarian populism, which include raging inflation, corruption, and
violent crime, are well known and act as a deterrent. Left-wing populism
and soft authoritarianism are unlikely to infect those countries.
Where leftists have achieved political success in those nations, they
differ profoundly from Chávez. With coherent organizations and agendas, the Brazilian Workers’ Party, Chile’s Concertación, and Uruguay’s
Broad Front have eschewed personalism and populism. Committed to
existing institutions and gradual change, they have preserved and en-
32
Journal of Democracy
riched democracy. Thus the authoritarian turn in Latin America today
comes not from the left in general, but from a populist left that in certain
countries is even more dangerous than its rightist forebear. The scrim
of “progressive” rhetoric around this undemocratic style of politics only
makes things worse.
Chávez’s death may abate this threat a bit, but competitive authoritarianism will likely persist and continue to hold appeal. The original Bolivarian leader is now gone, and Venezuelan subsidies may shrink, weakening especially Ortega in resource-poor Nicaragua. But the lessons of
Chávez’s remarkable “success” live on and may inspire more imitators,
particularly in Argentina. The undemocratic incumbents in Bolivia, Ecuador, and Venezuela have entrenched their rule and wield many tools
for extending it, aided by the commodities boom. Moreover, neither the
domestic opposition nor the international community has found a way to
stop discriminatory legalism. For these reasons, the end of the authoritarian trend in Latin America is not in sight.
NOTES
1. Scott Mainwaring, “The Surprising Resilience of Elected Governments,” Journal of
Democracy 10 (July 1999): 101–14.
2. For this concept, see Steven Levitsky and Lucan Way, Competitive Authoritarianism: Hybrid Regimes After the Cold War (Cambridge: Cambridge University Press, 2010).
3. Democracy’s replacement by competitive authoritarianism has yet to take on sharp
contours in every case. This suffocation is more evident in Venezuela under Chávez’s successor Nicolás Maduro, who has reacted to a spirited opposition challenge with ever more
repression, than in Ecuador, where a weak, fragmented opposition allowed Correa to win
an easy reelection victory in February 2013.
4. Steven Levitsky and James Loxton, “Populism and Competitive Authoritarianism
in the Latin America,” paper for 108th Annual Meeting, American Political Science Association, New Orleans, 30 August 2012; Raúl Madrid, The Rise of Ethnic Politics in Latin
America (Cambridge: Cambridge University Press, 2012), 178–83.
5. The longstanding debates on the concept of populism are reviewed in Kurt Weyland,
“Clarifying a Contested Concept: ‘Populism’ in the Study of Latin American Politics,”
Comparative Politics 34 (October): 1–22.
6. For a thorough analysis from an opposition standpoint, see Allan Brewer-Carías,
Dismantling Democracy in Venezuela: The Chávez Authoritarian Experiment (New York:
Cambridge University Press, 2010).
7. Although Morales has imitated Chávez’s populist script, Morales rose as the leader
of bottom-up social movements and therefore commands less personalistic autonomy and
control.
8. Fabrice Lehoucq, “Bolivia’s Constitutional Breakdown,” Journal of Democracy 19
(October 2008): 110–24.
9. Dexter Boniface, “The OAS versus ALBA: The Clash Over Democracy and the Rule
of Law in the Americas,” paper for 108th Annual Meeting, American Political Science Association, New Orleans, 30 August 2012.
Trustees of Princeton University
Rethinking Recent Democratization: Lessons from the Postcommunist Experience
Author(s): Valerie Bunce
Source: World Politics, Vol. 55, No. 2 (Jan., 2003), pp. 167-192
Published by: Cambridge University Press
Stable URL: http://www.jstor.org/stable/25054217 .
Accessed: 26/01/2015 09:10
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].
.
Cambridge University Press and Trustees of Princeton University are collaborating with JSTOR to digitize,
preserve and extend access to World Politics.
http://www.jstor.org
This content downloaded from 212.175.32.130 on Mon, 26 Jan 2015 09:10:26 AM
All use subject to JSTOR Terms and Conditions
RETHINKING
RECENT
DEMOCRATIZATION
Lessons
from the
Postcommunist
Experience
By VALERIE BUNCE*
Recent
Democratization
of recent democratization?of
such issues as
understanding
new
the
the
consolidation
of
democracies?has
been
OUR
origins and
of Latin America
and southern
influenced
heavily
by the experiences
as
is not surprising. The
third wave of democratization,
Europe.1 This
in
termed
southern
and
then
Samuel Huntington
it, began
Europe
to Latin America.2 Moreover,
the
oscilla
moved
given
political
quickly
in
tions of the region
American
Latin
they study, specialists
politics
were
to address
of regime transi
questions
unusually well positioned
a
two
tion. Finally,
the
these
of
experiences
combining
regions offered
a
contained
of countries,
large number
comparative
advantage. They
over the course of a decade
all of which
had redemocratized
virtually
a
some
in terms of history and cul
commonalities
and
half; they shared
of transi
ture; and yet they varied with respect to the timing and mode
that makes
and differences
such a mix of similarities
tion. It is precisely
for instructive
comparison.
The
Europe
and new
breakdown
between
to broaden
states
of state
1989
and Eastern
in the Soviet Union
rise of new regimes
and the subsequent
an
this region provide us with
opportunity
recent
I
of
democratization.3
By broadening,
socialism
and 1991
throughout
the discussion
*I thank Nancy Bermeo, George Breslauer, Michael McFaul,
Phil Roeder, Kathryn Stoner-Weiss,
reviewers for their comments on this article.
Sidney Tarrow, and the anonymous
1
Transi
See, especially, Guillermo
O'Donnell,
Philippe C. Schmitter, and Laurence Whitehead,
tions
Authoritarian
Rule, vols. 1-4 (Baltimore: Johns Hopkins University
Press, 1986).
from
2
in the Late Twentieth Century (Norman: University
The Third Wave: Democratization
Huntington,
of Oklahoma
Press, 1991).
3
in East
Subversion: Social Science and Democratization
See also M. Steven Fish, "Postcommunist
Europe and Eurasia," Slavic Review 58 (Winter 1999); Valerie Bunce, "Comparative Democratization:
Studies 33 (August-September
2000); Thomas
Big and Bounded Conclusions,"
Comparative Political
"The End of the Transition
Carothers,
Paradigm," Journal ofDemocracy 13 (January 2002).
WorldPolitics 55 (January2003), 167-92
This content downloaded from 212.175.32.130 on Mon, 26 Jan 2015 09:10:26 AM
All use subject to JSTOR Terms and Conditions
168
WORLD
POLITICS
to the
of the conversation.
If recent de
refer, most obviously,
geography
a
mocratization
then
the
terrain
of these
is, indeed,
global process,
studies should better reflect that fact. Moreover,
the
only by expanding
can
we
our
as
horizons
know whether
conceit
social sci
geographical
entists?that
of generalizability?is
well founded.
is, our presumption
There
three other aspects of broadening.
is the
One
are, in addition,
to
more
familiar
central
the
of
that
argument,
ideology
pluralism,
are
to fewer in
outcomes.
voices
is
This
preferable
quality
producing
in
the
be
correlation
comparative
particularly
important
politics, given
tween
and intellectual
As we all know, the
boundaries.
geographical
all tend
used, the questions
asked, and the theories evaluated
concepts
to take on a
cast.
regional
our fa
is a second consideration.
outside
Just as familiar
Stepping
terrain
miliar
often
alerts
us to new
factors
and new
relationships?
new
to borrow from Gorbachev.
As already
generally,
thinking,
a matter
not
is
of
this
intellectual
from
benefits
reaping
suggested,
just
a function
liberalization
of trade among
is
cultures.
This
also
scholarly
more
of the new
issues
that additional
the rise of new
with
the former
communist
states
world
ars to a series of
previously
pact of economic
regime
critical
most
distinction
and
the
often
economic
introduce.
For
came
sensitivity
heightened
concerns. These
overlooked
include
transition
on
example,
in
regimes
schol
among
and political
the im
the democratic
the
project;4
new democracies
(as in
founding
genuinely
versus
with
much
of
redemocratization
states)
(as
between
postcommunist
Latin America
cases
and new
and southern Europe); the impact of identity politics
state on democratization;5
the consequences
for democratic
politics of deficiencies in civil and political society;6 and the role of in
institutions
ternational
new democracies.7
in founding,
sustaining,
and/or
undermining
4
in the Postcommunist World," East
of Economic Reform
See M. Steven Fish, "The Determinants
European Politics and Societies 12 (Winter 1998).
5Juan Linz and Alfred Stepan, Problems ofDomestic Transition and Consolidation: Southern Europe,
South America, and Postcommunist Europe (Baltimore: Johns Hopkins University
Press, 1996); Debo
in Latin America,"
rah J. Yashar, "Democracy,
and the Postliberal Challenge
Indigenous Movements,
World Politics 52 (October 1999).
6
in Russia: Modern,
and Anti-Modern,"
Richard Rose, "Uses of Social Capital
Pre-Modern,
The Weakness of Civil Society in Post
Post-Soviet Affairs 16 (January-March
2000); Marc Howard,
Communist Europe (Cambridge: Cambridge University
Press, 2002).
7
See Daniel C. Thomas,
The Helsinki Effect: International Norms, Human Rights and the Demise of
Collision and Collusion: The
Communism
Press, 2001); Janine Wedel,
(Princeton: Princeton University
to Eastern Europe (New York: St. Martin's,
1998); Sarah L. Henderson,
Strange Case ofWestern Aid
Sector in Russia," Com
Aid and the Non-Governmental
Organization
"Selling Civil Society: Western
parative Political Studies 35 (March 2002).
This content downloaded from 212.175.32.130 on Mon, 26 Jan 2015 09:10:26 AM
All use subject to JSTOR Terms and Conditions
RETHINKING
DEMOCRATIZATION
A final benefit of broadening ismethodological.
169
The most illumi
are those
that restrain the universe
of causes while
comparisons
the range of results. In the case of comparative
democratiza
expanding
and
in
while
Latin
America
far
the
southern
tion,
go
meeting
Europe
are less
on the
recent
first condition,
they
helpful
second?though
threats to democracy
and perhaps
in, say, Peru, Venezuela,
Colombia,
nating
in
have provided
variables.
greater variation
Argentina
dependent
By
the
of
East-Central
and the
contrast,
postcommunist
region
Europe
former Soviet Union
is unusually
useful on both counts, given, for ex
ample, similarities in institutional legacies and in both the timing and
the agenda of transformation
the sheer diversity
of the re
alongside
and political
Charles King has aptly
gion's economic
pathways?what
variables" of
termed the "mercurial dependent
postcommunism.8
as an ideal
in
for comparative
laboratory
appeal of this region
are thus a number of studies
not been lost on
There
has
quiry
analysts.
area to address such
that use cases from the postcommunist
questions
The
as
either do or do not arise and why some of the new
why democracies
variations
others break down;9 whether
democracies
succeed, whereas
or
more
recent
in economic
reflect historical
influences
performance
or
transitions
economic,
factors;10 and why
political
geographical,
are sometimes
nationalist
protests, why
democracy
by
accompanied
some states dissolve
in reaction to these protests,
and why state dissolu
and
to
tion
is either
violent
or
peaceful.11
8
and the End of'Eastern
Transition,
Comparison,
Europe,'" World
King, "Post-Postcommunism:
Politics 53 (October 2000). On the divergent political and economic dynamics of the postcommunist
Slavic Review 58 (Winter 1999);
of Postsocialism,"
region, see Valerie Bunce, "The Political Economy
"Post-Communism's
Karen Dawisha,
Troubled
Steps toward
www.freedomhouse.org/ratings/index;
New States" (Manuscript, Cen
An Aggregate Analysis
of Progress in the Twenty-seven
Democracy:
ter for the Study of Post-Communist
of Maryland,
1997).
Societies, University
September
9
and Dictatorship:
Transitions
Michael McFaul,
"The Fourth Wave of Democracy
Noncooperative
of
in the Postcommunist
World," World Politics 54 (January 2002); M. Steven Fish, "The Dynamics
both in
Democratic
and Stephen E. Hanson,
Consolidation,"
Erosion,"
"Defining Democratic
Richard D. Anderson, M. Steven Fish, Stephen E. Hanson,
and Philip G. Roeder, eds., Postcommu
nism and the Theory ofDemocracy (Princeton: Princeton University
Press, 2001). Interwar Europe and
II Latin America
also provide an ideal laboratory for the study of democratic breakdown.
post-WW
See Nancy Bermeo, Ordinary People inExtraordinary Times: The Citizenry and the Collapse ofDemocracy
(Princeton: Princeton University Press, 2002).
10
to Rus
Vladimir Popov, The Political Economy of Growth in Russia, Program on New Approaches
sian Security, no. 17 (Cambridge: Davis Center, Harvard University,
2000); Timothy
Frye, "The Per
in the Postcommunist
ils of Polarization:
Economic
Performance
World," World Politics 54 (April
2002); Anders Aslund, Building Capitalism: The Transformation
of the Soviet Bloc (Cambridge: Cam
and
and David A. Reilly, "Geographical Diffusion
Press, 2002); Jeffrey S. Kopstein
bridge University
the Transformation
of the Postcommunist World," World Politics S3 (October 2000).
11
Mark Beissinger, Nationalist Mobilization
and the Collapse of the Soviet State (Cambridge: Cam
Press, 2002); Valerie Bunce, Subversive Institutions: The Design and theDestruction
of
bridge University
Socialism and the State (Cambridge: Cambridge University
Press, 1999); Carol Skalnik Leff, "Democ
inMultinational
States: The Breakup of the Communist
ratization and Disintegration
Federations,"
This content downloaded from 212.175.32.130 on Mon, 26 Jan 2015 09:10:26 AM
All use subject to JSTOR Terms and Conditions
WORLD POLITICS
170
Comparing
This
to use
aims
article
New Democracies
in East-Central
cases in all?to
Europe
rethink our understanding
of recent democratization.
It does so by con
a conversation
two bodies of research:
between
(1) studies of
ducting
Latin America
and southern Europe, which
have consti
collectively
and
the former
the postcommunist
experience
Soviet Union?twenty-seven
in the field, and (2) research on postcom
the reigning wisdom
will focus on two relationships
central
politics. The discussion
tuted
munist
to discussions
quent
in the field?between
regime
transitional
and between
trajectories
and subse
politics
the consolidation
and the sus
see that
will
We
the postcommunist
democracy.
issues
the
both
have
been
understood.
way
challenges
experience
In
I argue the
First, the degree of uncertainty
particular,
following.
of
tainability
transitions
democratic
varies
This
in turn
affects
in
the
considerably.
can
and their payoffs. Second, mass mobilization
to both the
and
the
consolidation
of
founding
democracy.
under certain conditions
is
furthered
the democratic
by
project
of transition
strategies
contribute
Third,
transitions
that
involve
boundaries.
both
nationalist
while
in state
and changes
protest
in democratic
consolidation
Fourth,
rapid progress
for
the
democratic
survival in the future, it does not
prospects
improves
are
less sustainable.
follow that unconsolidated
democracies
necessarily
to
contribute
Indeed,
(and the state) may
compromising
democracy
new
democra
democratic
survival. Finally, while
among
comparisons
cies can identify the optimal
conditions for democratization,
they may
have less to say about optimal
strategies for democratization.
Transitions
to Democracy:
Assumptions
and Arguments
on some core
premised
to
from dictatorship
about transitions
democracy?with
assumptions
an evident weak
as
with
the transitional
understood
beginning
period
elec
the
first
rule
and
with
of
authoritarian
competitive
ening
ending
The
analysis
tions. These
influences
of recent democratization
assumptions
are more
has been
include the following:
(1) that immediate
in shaping
important
un
are
transitions
that
(2)
dynamics;
inherently
quite
a
in
is
be
transition
the
central
that
(3)
bargaining
dynamic
than historical
considerations
transitional
certain;
Eurasia's
World Politics 51 (January 1999); Charles King, "The Benefits of Ethnic War: Understanding
States," World Politics 53 (July 2001); Abby Innes, Czechoslovakia: The Short Goodbye
Yale University
Press, 2001); Rasma Karklins, Ethnopolitics and the Transition toDemoc
Press, 1993).
racy (Baltimore: Johns Hopkins University
Unrecognized
(New Haven:
This content downloaded from 212.175.32.130 on Mon, 26 Jan 2015 09:10:26 AM
All use subject to JSTOR Terms and Conditions
RETHINKING
tween
with
authoritarian
outcomes
leaders
a function
issues on the table during
democratic
rule, building
authoritarians.12
These
assumptions,
and southern
America
about
what
and
leaders
of their
with
Europe,
the ideal
power;
opposition,
(4) that the key
with authoritarian
and
are
breaking
and eliciting
institutions,
have
constitutes
of the democratic
relative
the transition
coupled
171
DEMOCRATIZATION
the cooperation
of
studies
of Latin
comparative
several
generalizations
produced
approach
than thirty
as
First,
de
successful
to transition.
more
years ago,
argued
to
at the very least a
of the
prior settlement
require
state
about the rules of the
national
and
Second, bargaining
questions.13
to a small
transition
and the new political
order should be limited
Dankwart
mocratization
Rustow
seems
and representatives
of the democratic
op
to
it is useful
the
of
transitions,
position.
given
uncertainty
Finally,
construc
that promote
stability during the
forge compromises
political
this means pacting; reducing
the
tion of a democratic
order. In practice,
group
elites
of authoritarian
range of issues on the bargaining
of the state and, if possible, major
reforms); demobilizing
table, while
depriving
table
(for example,
and inherently
avoiding
destabilizing
reforms
economic
publics (which also limits the issues on the
the authoritarians
of a rationale
for sabotaging de
to
leaders agreeable
with
interim governments
mocratization);
forming
maneuver
some room for
in the
both sides; giving the military
political
a
a
that
and
election
constitution;
gov
holding
competitive
produces
ernment
and democrats.14
of both authoritarians
broadly representative
Mass Mobilization
seems
postcommunist
experience
about transitional
strategies.
sumptions
The
12
O'Donnell,
Latin America,"
to
of these
challenge many
Let us begin by addressing
as
the
in
andWhitehead
(fn. 1); Terry Lynn Karl, "Dilemmas of Democratization
Di Palma, To Craft Democracy
Politics 23 (Spring 1990); Guiseppe
Comparative
of California Press, 1990).
University
(Berkeley:
13
to Democracy:
Toward a Dynamic Model,"
Rustow, "Transitions
Comparative Politics 2 (April
1970).
14
of a
and Whitehead
O'Donnell,
(fn. 1); Richard G?nther,
Schmitter,
"Spain: The Very Model
and Richard G?nther,
Modern
Elite Settlement,"
in John Higley
eds., Elites and Democratic Consoli
dation in Latin America and Southern Europe (Baltimore: Johns Hopkins
Press, 1992);
University
to
Southern Europe's Transition
Robert M. Fishman,
State and Regime:
"Rethinking
Democracy,"
and Robert Kaufman, The Political Economy ofDem
World Politics 42 (April 1990); Stephen Haggard
Schmitter,
"Politics and
ocratic Transitions
Press, 1995); Jose Maria Maravall,
(Princeton: Princeton University
in Southern Europe," in Luiz Carlos Bresser Pereira, Jose Maria Maravall,
Policy: Economic Reforms
and Adam Przeworski,
eds., Economic Reforms inNew Democracies: A Social Democratic Approach (Cam
not imply the absence of
Press, 1993). Moderate
policies, however, do
bridge: Cambridge University
political conflict. See Nancy Bermeo, "Myths of Moderation,"
Comparative Politics 29 (April 1997).
This content downloaded from 212.175.32.130 on Mon, 26 Jan 2015 09:10:26 AM
All use subject to JSTOR Terms and Conditions
172
WORLD
POLITICS
role of mass
in the transition.
It is widely
spe
agreed among
publics
over time
and confirmed
the
Freedom
House
by
by
rankings
to
in the postcommu
transitions
that the most
successful
democracy
nist region have been in the Czech Republic,
Latvia,
Estonia, Hungary,
to
in
transition
and Slovenia.15 The
Lithuania,
Poland,
every
democracy
one of these cases, except
mass
Hungary,
began with
protests.16 More
im
that show significant
over, ifwe restrict our focus to those countries
cialists
over time, or
and
Bulgaria
performance
see
mass
same
at
mobilization
the
the
Romania,
pattern:
beginning
of the transition.
in their democratic
provement
we
Why
sition
was mass mobilization
so often
in the postcommunist
a number
protests
performed
of the authoritarian
breakdown
there were
alternatives
sometimes
even
leaders
to the democratic
tran
helpful
answer
context? The
is that political
of valuable
the
functions. They
signaled
order;
created
a
sense
widespread
authoritarian
to that order;
pushed
as with Walesa
of the opposition,
leaders
in Poland)
that
(and
to
the bargaining
restored) a large opposi
table; created
(and sometimes
tion united by its rejection of the incumbent
regime; and gave opposi
authoritarian
tion leaders a resource advantage when
with
bargaining
for radical change
created a mandate
elites. Finally, mass mobilization
translated
that
subsequently
forces in the first
introduction
competitive
of far-reaching
into
a
for the democratic
large victory
that, led to the
and, following
economic
and political
reforms.
elections
Uncertainty
can further
mass mobilization
the transition
accept that
during
chal
confront
additional
then we necessarily
the democratic
project,
recent democratization.
to the received wisdom
it
about
First,
lenges
can be
in the postcommu
in many cases such mobilization
that
argued
a
of the transition?by
nist region reduced
the uncertainty
providing
the bargaining
of mass
clear reading
sentiments,
by strengthening
If we
power
of opposition
leaders,
and by forcing
the communists
to
give up
their defense of the old order, either stepping aside quickly (as in
Czechoslovakia)
or, when
thinking
prospectively,
joining
the movement
15
See fn. 8.
16
out ultimately
was understood
to be politically
In Hungary mass mobilization
risky (and turned
in 1956, on the one
Revolution
to be unnecessary),
given the brutal suppression of the Hungarian
to jump on the
even before the roundtable,
of the reform communists,
hand, and the willingness
on the other hand. See Patrick H. O'Neil,
Insti
"Revolution
from Within:
democratic
bandwagon,
and the Case of Hungary," World Politics 48
tutional Analysis, Transitions
from Authoritarianism,
(July 1996).
This content downloaded from 212.175.32.130 on Mon, 26 Jan 2015 09:10:26 AM
All use subject to JSTOR Terms and Conditions
RETHINKING
173
DEMOCRATIZATION
for democracy (as in Poland, Slovenia, and the Baltic states).17At the
same
delivered?a
time, mass mobilization
promised?and
popular
in the first competitive
for democracy
elections.
to
Most
in the postcommunist
of the transitions
world
democracy
is evidenced
the
that
the
fact
were, of course, highly uncertain. This
by
mandate
in most of the countries
in the region led to a
election
competitive
au
more
communist
the
the
the
victory. Indeed,
larger
victory,
likely that
even ten years after the transition
thoritarian
rule continued. Moreover,
were ranked
one-third
of
the
began, only
postcommunist
regimes
fully
first
this is the highest number
free. Although
since state socialism fell, it is a
one
in the
finds at a comparable
lower than what
percentage much
point
Latin American
and southern European
transitions. When
combined
with
the earlier observations,
these patterns
suggest that the uncertainty
In
trajectories varied significantly.18
postcommunist
surrounding
political
some cases, a democratic
outcome was
in most oth
relatively predictable;
were far more
communism
after
ers, the political
options
open-ended.
Strategic
The
existence
of amore
Implications
certain
political
of Uncertainty
environment
in some countries
into question
both the necessity
and the logic, outlined
earlier, of
new
au
between
the
democracy
by forging compromises
safeguarding
to do
and democrats.
It is precisely
the absence of pressure
thoritarians
calls
so in the Polish,
transitions
the Czech,
and the other highly
successful
that explains another contrast between
the "East" and the "South." It is
true that many of the most
in the postcommunist
successful
transitions
as elaborate as the
area included
experi
pacting
Spanish
(though rarely
for a brief time broadly representa
ence) and that some also evidenced
tive interim governments.19
that the transitions
It is also true, however,
in the
that
combined
region
postcommunist
pacting with demobilized
publics?or
South?were
authoritarian
to be the
in the
preferred
approach
were most
to continue
the
transitions
that
likely
precisely
the other
rule in the postcommunist
region.20 Moreover,
what
has been
asserted
17
See Anna M. Grzymala-Busse,
the Communist Past: The Regeneration
Redeeming
of Communist
Parties inEast-Central
Press, 2002).
Cambridge University
(Cambridge:
Europe
18
Because Poland was the first country in the region to break with communist party rule, its transi
in
tion was somewhat more uncertain. Given
the character of the Soviet bloc, however, developments
of the bloc.
Poland during the first half of 1989 lowered the risks of transition for other members
of
19Jon Elster, ed., The Roundtable Talks and the Breakdown of Communism
(Chicago: University
Press, 1996).
Chicago
20
Pauline Jones Luong, Institutional
Power, Perceptions and Pacts (Cambridge:
in Post-Soviet
Change and Political Continuity
Press, 2002).
Cambridge University
This content downloaded from 212.175.32.130 on Mon, 26 Jan 2015 09:10:26 AM
All use subject to JSTOR Terms and Conditions
Central Asia:
WORLD POLITICS
174
that
compromises
and Latin American
were
so beneficial
deemed
for the southern
European
leaders in
rejected by opposition
were
the like. Instead,
they
strongly
the authoritar
and sharp break with
were
transitions
and
Slovenia,
Hungary,
to favor an immediate
positioned
Poland,
case of democratization
in the
in every highly
ian past. Thus,
successful
was excluded
from
from
the
influence
the
region,
military
political
a
with
the
the
first
elections
involved
radical
break
lead
start;
political
were
in
the
introduced
the
and
of
economy
past;
major changes
ership
as
was the commitment
in each of these cases
important
quickly.21 Just
to
most
in
its very boundaries.
the
of
them
state, including
reforming
For the Czech Republic, Hungary, Poland, Slovenia, and the Baltic
ambitious.
states, then, the agenda of transition was unusually
transition dynamics
therefore ask us to amend the fa
Postcommunist
drawn from the South. It was precisely because mass
miliar formulation
was so
to authoritarians
that leaders of the op
threatening
some of these countries were free to carry out radical
in
political
position
because of popular mobilization
and economic
reforms. Put differently:
mobilization
case, reform
or, in the Hungarian
tween democrats
and authoritarians,
sustainable
and full-scale
the most
communism
opposition
democracies
and collaboration
leaders
in what
in the East
be
became
could
pro
ceed quickly in breaking with authoritarian rule and building democratic
(and, for that matter,
as their counterparts
in turn,
This,
capitalist)
elsewhere
altered
without
as much
appeasing
of transition
interests.
institutions
about
the strategies
worrying
authoritarian
and
their payoffs.
bridging between the old and the new order constituted by all
While
accounts
America
communist
the most
successful
and southern
region
was
approach
the most
Europe,
the opposite?severing
The Role
to democratization
successful
strategy
in Latin
in the post
ties.
of the Military
strate
contrasts
in the optimal
interregional
in
Latin
of
the
role
the
different
very
military
gies
one
on
in
communist
and
the
the
America
and southern Europe,
hand,
es
in the South have argued with
area, on the other hand. Specialists
as in the
to democracy
one voice that the
threat
today,
biggest
sentially
Also
contributing
of transition
to these
was
21
On the benefits of breaking with past leadership, see Aslund
(fn. 9);
(fn. 10); Frye (fn. 10);McFaul
an
reforms were introduced
economic
and Bunce (fn. 8). Hungary,
exception. Major
again, provides
returned to power with a large
the ex-communists
election, when
only after the second competitive
in how far reforms went prior to the end of com
mandate. However, Hungary was also exceptional
munist party rule.
This content downloaded from 212.175.32.130 on Mon, 26 Jan 2015 09:10:26 AM
All use subject to JSTOR Terms and Conditions
RETHINKING
past,
tory
which
DEMOCRATIZATION
175
is the military. One has only to recall, for
the long his
example,
most
in Latin American
of military
interventions
of
politics,
a return
some of them oversaw
democracy
(though
as
in
circuitous
also
occurred
fash
governance,
unusually
terminated
to democratic
ion, in the Portuguese
transition).
There
is, in addition,
the attempted
military coup d'?tat in Spain in 1982. Indeed, precisely because of its
in
the military
has been awarded remarkable
politics,
their democratic
claims
Latin American
constitutions,
an
When
these
obvious
combined,
carry
examples
notwithstanding.22
can make or break
in these contexts
the military
It is
message:
regimes.
long
importance
in many
powers
this capacity that contributed
precisely
tions in the South and that necessitated
to the
of the transi
uncertainty
with
authoritar
compromises
ian forces.
In much
tradition
of the postcommunist
of civilian control over
there is a long
world,
by contrast,
the military?a
tradition
that goes far
back in Russian history and that, following the Bolshevik Revolution
and
the demilitarization
after
the Civil War,
was maintained
at home
II was
to the members
War
outward
of
projected
one
area
in
bloc.23 Civil-military
constituted
relations,
short,
the authoritarian
rather than a bur
where
past proved to be beneficial,
state
after
socialism.24
den, for democratization
context and
in the postcommunist
the military
less threatening
With
au
in support of democracy,
in some cases mobilized
with mass publics
and then after World
the Soviet
were
in the postcommunist
indeed under siege.
region
the case in East-Central
This was particularly
Europe, where domestic
secret
in Yugoslavia,
control over the military
the
(and
police)?except
to
the
Union
Soviet
after
and
Albania?had
been
ceded
Romania,
thoritarian
elites
success
came to be the most
in what
this left the opposition
in the
freedom
of maneuver?a
ful democracies
region with unusual
1968. All
a result, both the
support in the streets. As
tran
most
and the
successful
effects
strategies of
context from what
in the postcommunist
sition were different
they had
been in Latin America
and southern Europe.
freedom
enhanced
of mass
by public
mobilization
22
in
Political Transitions
and Military
Brian Loveman,
"Protected Democracies
Guardianship:
Studies and World Affairs 36 (Summer 1994).
Latin America,
1978-1993," Journal ofInter-American
23
or
The key phrase is "members of the Soviet bloc." For those communist
regimes outside the bloc
mavericks within
the bloc (Albania, Romania,
and Yugoslavia),
party control over the military was
It was precisely in these cases that the exit from state socialism was violent. Variations
compromised.
of the Yugoslav
in civil-military
relations also account in part for the violent disintegration
state, in
contrast to the
See Bunce (fn. 11).
of both the Soviet Union and Czechoslovakia.
peaceful dissolution
24
of
Market
"Rival Views of Postcommunist
See Bela Greskovits,
Society: The Path Dependence
inMichel Dobry, ed., Democratic and Capitalist Transitions inEastern Europe: Lessons for
Transitology,"
the Social Sciences (Dordrecht: Kluwer Academic
Publishers,
2000).
This content downloaded from 212.175.32.130 on Mon, 26 Jan 2015 09:10:26 AM
All use subject to JSTOR Terms and Conditions
WORLD
176
POLITICS
Nationalist
Mobilization
an issue of considerable
thus far has sidestepped
analysis
tance in the transitions
from state socialism:
the distinction
The
protests
postcommunist
ular protest in both the Czech
the Baltic
liberal
and Slovene
and nationalist.
democratic
state. Here,
the
pop
pattern. While
the regime and protests
against
exhibits
another
region
surprising
against
governance,
impor
between
the
lands and Poland targeted the regime,
are better
demonstrations
understood
In the latter cases, then, nationalism
even when nationalist
concerns grew
as both
supported
out of and
were
in part responsible
for the disintegration
of a state.
nationalist
also seems to be another positive
linkage between
democratization.
The
mobilization
and sucessful,
sustained
republics
varied
that made up the Soviet Union, Yugoslavia,
and Czechoslovakia
There
from each other with respect to whether
publics protested,
considerably
was
the
whether
the opposition
strong and united, and whether
publics,
were
even the communists
to de
committed
and,
indeed,
opposition,
states
the breakup
mocratization.
of these three ethnofederal
With
the best conditions
for
lines, those republics with
along republican
were
from a political
and economic
liberated
democratic
governance
not
context
if
that made
such an outcome
impossible. Thus,
unlikely,
not
but also Macedonia,
and the Baltic
just Slovenia
republics,
to pursue a dem
and Ukraine were better positioned
Russia,
Moldova,
course
ocratic
can we
How
that
state
following
reconcile
nationalist
disintegration.25
with
these observations
mobilization
a threat
poses
the familiar
to
democracy
and democratization
that the logics of state building
has empirical
tradictory? This argument, moreover,
communist
the
deleterious
effects
world,
given
grounds
after
political
developments
and
Serbia
Montenegro
gia,
these
within
who
cases
the nationalist
state
socialism
the republic;
transformed
to maintain
then used nationalism
structed
illiberal
states.26 What
successor
explains
regimes
these divergent
are con
support in the post
on
of nationalism
in Bosnia, Croatia, Geor
In each of
and Slovakia.
(and Kosovo),
movement
excluded
some
argument
on the
communists
authoritarian
minorities
residing
into nationalists,
control; and con
successor
deconstructing
of
nationalism?
consequences
while
25
in Germany: A Reappraisal,"
"Democratization
For a parallel situation, see Michael
Bernhard,
no. 4 (2001).
Comparative Politics 33,
26
over
A Comparative
See Zsuzsa Czergo,
Study of Contestations
"Language and Democracy:
and Slovakia" (Ph.D. diss., George Washington
2000); Georgi
University,
Language Use in Romania
in
and Adjaria before and since the Soviet Collapse,"
"The Tale of Two Resorts: Abkhazia
Derluguian,
and Ronnie D. Lipschutz,
eds., The "Myth" ofEthnic Conflict: Politics, Economics and
Beverly Crawford
no. 98.
and
of California
International
Cultural Violence, Research Monograph,
(Berkeley: University
This content downloaded from 212.175.32.130 on Mon, 26 Jan 2015 09:10:26 AM
All use subject to JSTOR Terms and Conditions
RETHINKING
When
177
enters
nationalism
to thick
often
DEMOCRATIZATION
the discussion,
arguments
parsimonious
In
this
there
instance,
however,
explanations.
when
distinction:
nationalist
demonstra
simple
give way
to be a
relatively
tions began in the republics.
seems
alist demonstrations
Late
nationalist
that first
when
mobilization?or
nation
the communist
appeared
regime
in virtually
associated
every instance
disintegrating?is
a
to
with
since that time in
and progress
democracy
rapid transition
a stable?or
at least
stable?democratic
order.
building
increasingly
not
cases of Estonia,
in
This
the
describes,
Latvia,
just
particular,
and Slovenia,
but also the far more flawed, but nonetheless
Lithuania,
and
state were
durable
democracies
of Moldova,
and Ukraine.
Russia,
that first occurred before the
nationalist
demonstrations
contrast,
By
state
to unravel are associated with very different po
and
regime
began
or
litical pathways
after state socialism?either
democratic
breakdown
a
to
transition
and one
There were five republics
democracy.
delayed
autonomous
province
that experienced
such demonstrations
by their
titular nation during the 1970s or at the beginning of the 1980s: Ar
limited ex
Croatia, Georgia,
Slovakia,
Kosovo,
and, to a more
to
transition
tent, Serbia.27 In every one of these cases the subsequent
was undermined,
as was the successor
cases.
state in most
democracy
menia,
is timing so important? The key seems to be differences
Why
in
arose in re
In the "early" cases, nationalist mobilization
on
a
sense
the
of
conditions:
sponse
strong
part of
identity
members
and
the
leaders
of
the
tit
especially
self-appointed
republic's
ular nation
earlier developments,
such as the experience
of
(reflecting
to communist
statehood
with
po
party rule) coupled
prior
republican
litical dynamics
that featured
domination
by the titular nation
along
regime
context.
to two
from
the center. Once
demonstrations
autonomy
significant
minorities
within
three
followed:
these republics
began,
developments
themselves
from titular
defended
Armenia)
(except homogeneous
countermovements
the cen
domination
while
by building
allying with
with
ter;
the
thereby
that nationalist
would
protests
fearing
spread and
both
the
and
the
the titular
state, suppressed
challenge
regime
center,
Area
2d ed. (Bloomington:
Indiana
Studies, 1998); Ronald Suny, The Making
of the Georgian Nation,
and Sustaining De
Press, 1994); Valerie Bunce and Stephen Watts,
University
"Managing Diversity
versus Unitary States in the Postsocialist World"
at the work
mocracy: Ethnofederal
(Paper presented
on
and Peace-Making,
San Diego, December
10-11, 2001).
shop
Power-Sharing
27
See Beissinger
and State Disintegration:
The
(fn. 11); and Besnik Pula, "Contested Sovereignty
Rise of the Albanian
in Kosovo"
Secessionist Movement
(Master's Thesis, Georgetown
University,
"Liberalism and Minorities:
Serbs as Agents
and
2001). On the Serbian case, see Valere P. Gagnon,
on
inMulti
of Space" (Paper presented at the workshop
Victims of Liberal Conceptions
Citizenship
cultural States: Comparing
Institute of International Af
the Former Yugoslavia
and Israel, Austrian
fairs, Vienna,
April
20-21,
2001).
This content downloaded from 212.175.32.130 on Mon, 26 Jan 2015 09:10:26 AM
All use subject to JSTOR Terms and Conditions
WORLD
178
POLITICS
national
the republican
mi
party, and empowered
protesters,
purged
as a
to the titular nation;
norities
and the republican
counterweight
in
the
face
fissured
irreconcilable
of
demands
from
local national
party
ists versus central communists.
a result,
state socialism
to dissolve,
the stage
by the time
began
was
set for an
to both demo
transition
already
unusually
problematic
cratic rule and
statehood.
Two
insurmountable
divides
independent
As
were
was between
who dominated
the
nationalists,
place. The first
been
and
who
had
demobilized.
The
second
scene,
liberals,
political
was between
leaders of the majority
nation
and leaders of minority
communities.
of these groups were well defined
The national
identities
in
and their
and exclusivist,
interests,
peting
political
the communist
Moreover,
com
identities were joined with
competing
for
and
the
future.
alliances,
preferences
leaders of these republics,
facing the loss of
both their institutional and their ideological bases for ruling, did not
have
the option
their Slovenian
that
embraced
both
opposition
counterparts
to an
had, of defecting
statehood
and liberal de
independent
a
or, if adopting
Instead, they could either become nationalists
mocracy.
liberal position,
face political marginalization.
in re
when
nationalist
mobilization
By contrast,
began only later,
to
the
of
the
and
the
all
these
condi
state,
sponse
regime
weakening
tions were
absent?or
and
majority
at least
the minorities
were
less well
free
defined.
This
to coalesce
and liberal democracy. Thus,
sovereignty
republican
liberal agenda combined with a nationalist
agenda;
sition
We
communists
but even many
our discussion
can now conclude
forces
meant
the
around
embraced
of transitions
in these
of
contexts
a
and not only
that agenda.
in the South
sus the East. The
the
that
issues
oppo
ver
the following,
America
and
of the latter region suggest
experiences
counter
to the received wisdom
about Latin
all running
are
historical
factors
critical in shaping the re
southern Europe.
First,
as
sources and
the preferences
of elites during the transition,
especially
one
transition
as, more generally,
Second,
trajectories.
proximate
the transition,
of
outside
the
and positive
influence,
high politics
lying
seem to vary in their
is mass mobilization.
transitions
Third,
degree of
most
constitutes
successful
the
and this affects what
path.
uncertainty,
some transitions were
In the postcommunist
less uncer
world, where
on both
was one that moved
successful
tain, the most
quickly
approach
can
success
be
democratization
economic
fronts.
and
Fourth,
political
well
fulwhen it is combined with nationalist mobilization
of a new
gins with
and the founding
so when
is particularly
such mobilization
of the state and the regime.
the weakening
state. This
This content downloaded from 212.175.32.130 on Mon, 26 Jan 2015 09:10:26 AM
All use subject to JSTOR Terms and Conditions
first be
RETHINKING
179
DEMOCRATIZATION
Finally, ifwe divide the transitions in the postcommunist world into
two
was
nationalist
types?where
not?we
find
two
was present
and where
has already been noted?the
of timing?when
nationalist
mobi
mobilization
it
stories. One
simple
for democratization
consequences
lization begins. The
second story describes
in
the remaining
countries
the region. Here,
the key issue appears to be the strength of the oppo
in the first election.
Put
sition, as indicated
by their competitiveness
more
the better their electoral performance,
the
successful
succinctly,
to
the transition
Democratic
With
democracy.
the formation
Sustainability
elected government,
of the first popularly
to have ended. Scholars
is understood
to
sition
and Democratic
Consolidation
democracy
their focus to two
the tran
then
shift
of democracy
and its sus
to the
to
of
consolidation
refers
The
democracy
degree
tainability.
which
the key elements
of a democratic
order are in place, and whether
those
elements
issues:
the consolidation
to promote
function
effective,
and account
inclusive,
able governance.
of democracy
The
refers, simply, to the
sustainability
continuation
rule. In the first case, scholars look to mass
of democratic
interest associations,
and political
participa
(public opinion,
publics
institutions
the state, and representative
tion), political
bodies),
(parties,
and the behavior
In the second,
the concern
of
of political
leaders.
has been
investigations
scholarly
equally
broad,
looking
to economic
and demographic factors, political institutions, political parties, public
of political
leaders.
one another, which
is
confront
rarely
since on the face of it they are analytically
related. Despite
surprising,
two lines of
seem to converge
the parallel play, however,
do
along
they
have pow
is
that
made
leaders
the
choices
One
argument.
by political
on
it sur
consolidates
and whether
whether
erful effects
democracy
opinion
These
and behavior,
two bodies
vives.28 The
to which
other
and the decisions
of work
is that the quality
as defined
of democracy?that
it is consolidated
Russia's Muddied
above?predicts
Political
is, the degree
its sustainability.
Profile
to
two arguments
than
that Iwish
question. Rather
as
cases
I did earlier, I
with
the postcommunist
each other,
comparing
serves
will concentrate
instead on one case in particular: Russia, which
It is precisely
these
28
See, for example, Higley
and G?nther
(fn. 14).
This content downloaded from 212.175.32.130 on Mon, 26 Jan 2015 09:10:26 AM
All use subject to JSTOR Terms and Conditions
180
WORLD
POLITICS
as a useful
is that Russia
foil for two reasons. One
represents
empirical
re
in many
the
modal
case?for
respects
postcommunist
example, with
to
economic
and
economic
the
spect
age of
development
performance,
the state, the structure of its government,
of labor, and the
the weakness
the most
system.29 Perhaps
surprising
one that,
the central
politics?and
again, represents
in postcommunist
the absence of significant
Eurasia?is
po
tendency
litical polarization
and the relative stability of their po
among citizens
slow development
aspect of Russian
of the party
over time.30 The other reason to focus on Russia
is
preferences
a
test of the arguments
the case provides
about
particularly
good
consolidation
and sustainability. This
is because Russia has a highly
im
litical
that
probable and seriously flawed, yet durable democracy. The
can be summarized
Russian
as follows.
Russian
democracy
in two key respects. One
is the Russian
problem
a second is the weakness
state. As is
of the Russian
widely
is deficient
ledger of
democracy
presidency;
recognized,
a central role in the rise of
in Russia,
played
democracy
seem
to
in
of
his
would
have com
1993,
actions, beginning
yet many
as well as economic
the democratic
reform and state
promised
project,
to
his
commitment
the
recentralization
of
given
capacity.31 Moreover,
the Russian
Yeltsins
Vladimir
could
be
viewed
state,
successor,
Putin,
Boris
Yelstin
as a less
capricious
democratic
against
a more formidable
force
therefore, potentially
an
is
Such
interpretation
particularly
politics.
Russia
and
the parallels
between
contemporary
and,
tempting,
given
Weimar
Germany?for
downward
mobility
both
cases
tant powers
economic
disastrous
performance,
example,
in
in the international
system, and the existence
of a mixed
with
system,
impor
presidential-parliamentary
reserved
for the presidency.32
29
"The Limited Reach of Russia's Party
See, for example, Bunce (fn. 8); Kathryn Stoner-Weiss,
System," Politics and Society 29 (September 2001); Stephen Crowley and David Ost, eds., Workers after
Workers' States: Labor and Politics in Post-Communist
and Littlefield,
Europe (Lanham, Md.: Rowman
2001).
30
in Russia," Communist and Postcommunist
See, especially, Richard Ahl, "Society and Transition
in
of Mass Partisanship
Studies 32 (April 1999); and Ted Brader and Joshua Tucker, "The Emergence
45 (January 2001); for a different view, see Valerii Solovei,
Russia,"American
JournalofPoliticalScience
in Lilia Shevtsova,
i natsionalisticheskoi
transformatsii Rossii,"
ed., Rossiia:
"Kommunisticheskaya
(Russia: Politics) (Moscow: Carnegie Center, 1998).
politichestkaya
31
Three very insightful analyses of the Yelstin era are Lilia Shevtsova, Yeltsins Russia: Myths and Re
alities (Washington, D.C.: Brookings
Institution,
1999); Michael McFaul, Russia's Unfinished Revolu
tion: Political Change from Gorbachev to Yeltsin (Ithaca, N.Y.: Cornell University
Press, 2001); and
as Leaders
Press, 2002).
George Breslauer, Gorbachev and Yeltsin
(Cambridge: Cambridge University
32
Post-Soviet
See Stephen E. Hanson
and Jeffrey S. Kopstein,
"The Weimar/Russia
Comparison,"
is one striking contrast between
these two cases:
13
there
1997).
However,
(July-September
Affairs
were
civil society in interwar Germany was highly developed but German
polarized, whereas
publics
civil society in Russia is far less developed but Russian publics tend to cluster at the center of the po
This content downloaded from 212.175.32.130 on Mon, 26 Jan 2015 09:10:26 AM
All use subject to JSTOR Terms and Conditions
RETHINKING
181
DEMOCRATIZATION
is a
of the
so, too, is the weakness
presidency
problem,
state. It isweak because of the absence of rule of law, continu
Russian
central and local laws, and the privileging
conflicts
between
of lo
ing
If the Russian
over
calities
the
center
with
to both
respect
the
identities
of
local
and publics
and the targets of compliance.
officials
As a consequence,
are both
state and
the Russian
regime
spatially fragmented.33 This frag
to
extends
the
moreover,
mentation,
economy,
thereby compromising
the capacity of the state to create the economic
necessary
integration
to function
for capitalism
effectively.34
If the internal boundaries
of the Russian
then
the external
boundaries
of this
state
Federation
are "too
strong,"
is evi
This
are "too weak."
wars in
in the continuing
Also
dent, for example,
important
Chechnya.
the
is the absence
boundaries
of the
of public agreement
concerning
in practices
and in
state?which
reflects both the power of localities
of Russian
and the weakness
iden
national
public and official minds
is
Russian
the
absence
weak
for
reasons,
many
tity.
identity
including
the Soviet period
institutions
that could forge
of republican
during
such an identity; the divisions within the Soviet and then Russian elite
over whether
stratum
able constraints
on
to define
this identity, and the consider
the
of a common
identity when
in question
is numerically
dominant,
geograph
an
in the core, not the periphery, within
located
and how
the construction
protonation
and
ically dispersed,
a state.35 Thus,
not within
lacked
Russian
national
identity
empire,
the former Soviet Union
what many other groups within
had and what
particular
served
to
as the basis for mobilization
disintegrate:
ographical
political,
compactness;
once
the regime and the state began
institutions;
cultural, social, and economic
ge
an "other" that was defined
in
simultaneously
Civil Society, and the Prospects for Consoli
litical continuum.
See James Gibson,
"Social Networks,
Science 45 (January 2001); and
Tradition," American Journal ofPolitical
dating Russia's Democratic
Sheri Berman,
"Civil Society and the Collapse
of theWeimar
Republic," World Politics 49 (April
1997).
33
in Post-Soviet
Russia," Post-Soviet Af
See, in particular, Kathryn Hendley,
"Legal Development
and Retrench
13 (October-December
1997); Kathryn Stoner-Weiss,
Resisting the State: Reform
fairs
ment in Post-Soviet
A historical
Russia
Press, forthcoming).
University
(Cambridge:
Cambridge
on rule of law in Russia
is provided
in George Yaney, "Law, Society, and the Domestic
perspective
in Russia inHistorical
Perspective," American Political Science Review 59 (June 1965).
Regime
34
See David Woodruff, Money Unmade: Barter and theFate ofRussian Capitalism (Ithaca, N.Y: Cor
nell University
Press, 1999).
35
and the Soviet State,
Russia: Russian Nationalism
See, especially, Yitzhak
Brudny, Reinventing
1953-1991
Press, 1998); and idem, "National Identity and Democ
(Cambridge: Harvard University
of
of Political Science, Hebrew University
Russia" (Manuscript, Department
racy in Postcommunist
and
2001). See also Veljko Vujacic, "Historical Capacities, Nationalist Mobilization
in Russia and Serbia: A Weberian
View," Theory and Society 25 (December 1996);
of Soviet
and the Deconstruction
Mark Beissinger,
"Demise of an Empire-State:
Identity, Legitimacy
ofWis
Politics," in Crawford Young, ed., The Rising Tide of Cultural Pluralism (Madison: University
Jerusalem, August
Political Outcomes
consin Press,
1993); Bunce
(fn. 11), chaps. 4-5.
This content downloaded from 212.175.32.130 on Mon, 26 Jan 2015 09:10:26 AM
All use subject to JSTOR Terms and Conditions
182
WORLD
ideological,
spatial,
and national
POLITICS
terms;
and a sense of being
an embat
tled minority poised against amajority.
If there
consensus
is a consensus
state
that the Russian
a serious
that this poses
has observed:
problem
Holmes
for Russian
Russia
Stephen
that liberal values
isweak,
there
is also a
democracy.
As
makes
"Today's
excrutiatingly
are threatened
as
state inca
just
thoroughly
by
as
most definitions
While
of
do
by despotic
pacity
democracy
power."36
not mention
com
the state and instead emphasize
and
liberties,
rights,
a
state is nonetheless
petition,
capable
implied. Can there be civil liber
plain
ties and political rights without rule of law?Can political competition
function
as it should
ing in the hands
in a democracy
of elected officials
for example, power
without,
state guarantees
and without
rest
that
there will be free and fair elections and that public policies will be
implemented?37
The Case for Russian Democracy
is seriously
that Russian
doubt,
therefore,
democracy
In some important
and unconsolidated.
de
respects, however,
is
in
Russia
has
well
since
defined
Russia.
First,
mocracy
independence,
more
at the national
at the local
held five elections
level?and
hundreds
There
is little
flawed
and regional levels. These
elections have by and large been free and fair:
and power has changed
have
invited
considerable
they
competition,
In the gubernatorial
hands
of 1996-97,
for ex
elections
repeatedly.
candidates were defeated?an
of the incumbent
fully one-half
one
is
since
indicator
that a democracy
consideration,
important
turnover of
the
rules
is
consolidating
peaceful
political power.38 Second,
stable since the referendum
of the political
have
been
game
relatively
ample,
on the 1993 Constitution. Third, the number of parties on the ballot
for the Duma
has
declined
over
time,
as has
the number
of wasted
over
time be
of growing
there is evidence
Fourth,
cooperation
are
some
ex
while
there
the president
and the Duma.40
Fifth,
as
in the Duma,
tends
tremists
Russian
already noted,
public opinion,
to cluster at the center of the
im
continuum
and to evidence
political
votes.39
tween
36
"What Russia Teaches Us Now" American Prospect 32 (July-August
1997), 33.
Holmes,
37
and the Rule of Law," American
"The Political Foundations
of Democracy
See Barry Weingast,
Political Science Review 91 (June 1997).
38
no. 115
Steven Solnick, Putin and the Provinces, Research Monograph,
(Cambridge: Harvard
on New
to Russian Security, 2000).
Approaches
University, Davis Center, Program
39
in
See Robert Moser, Unexpected Outcomes: Electoral Systems, Political Parties, and Representation
of Pittsburgh Press, 2000).
Russia (Pittsburgh, Pa.: University
40
in a Transitional Regime,
Thomas
The Russian Parliament: Institutional Evolution
Remington,
1989-1999
Press, 2001).
(New Haven: Yale University
This content downloaded from 212.175.32.130 on Mon, 26 Jan 2015 09:10:26 AM
All use subject to JSTOR Terms and Conditions
RETHINKING
183
DEMOCRATIZATION
pressively high support for democracy despite the uncertainty of the
environment.41
Sixth, the Russian
even
it
indeed,
may become
relatively well;
recent reforms are
if President
Putins
fully
our purposes,
most
for
importantl
gloomy
Russian
has lasted.
democracy
political
court
more
system has functioned
in the future
effective
implemented.42
predictions
and
Finally,
to the contrary,
It is precisely this final point that introduces two key questions. Most
obviously:
given the ledger just presented, why has Russian
less obviously:
is this in spite of its democratic
endured? And
or because of them?
is usually
presumed,
Comparative
democracy
deficits, as
Standards
can
to answer these
the issue of com
questions
begin
by addressing
for his role in weakening
has been criticized
standards. Yeltsin
parative
with
various
bilateral
treaties
the state (through,
say,
regional gover
We
the accumulation
nors),
tolerating
economic
interests, compromising
of considerable
economic
reform
power by various
in ways
that serve
a
that does not
interests, and, therefore, giving Russia
capitalism
to ex
in its
work.
It is a state that is both corrupt and limited
ability
meet
tract resources, elicit
its
and
and
financial obligations;
compliance,
those
it is a democracy
that
lacks
is that Yeltsin
these
in
is often
What
accountability.
implied
that sup
should have made other choices
and capitalism,
rather than undermining
critiques
the state, democracy,
ported
a
Were
ef
those superior choices
them. But this introduces
question.
to Yeltsin? The answer to that question
is rarely pro
available
fectively
or the Czech
another
Poland
vided.
Instead,
case?say,
Republic?is
to
costs
to demonstrate
of the Russian
the
introduced
approach
politi
cal and economic
There
Czech
are
transition.
good
reasons
or the
either Poland
pair Russia with
at
in
shock
is
variations
issue
hand
when
the
to
Republic,
especially
or voucher
the problem
However,
therapy
privatization.43
once
the
ond step that is sometimes
taken,
comparison
iswith
a sec
is raised:
to
41
"Tolerance, Transition
See, especially, Donna Bahry, Cynthia Boaz, and Stacy Burnett Gordon,
in Russia," Comparative Political Studies 30 (August 1997); James L.
and Support for Civil Liberties
Post-Soviet Affairs 17 (April-June 2001).
"The Russian Dance with Democracy,"
Gibson,
42
Herman
Schwartz, The Struggle for Constitutional Justice inPost-Communist Europe (^Chicago:Uni
in Russia," Post-Soviet
versity of Chicago Press, 2000); Robert Sharlet, "Putin and the Politics of Law
"Putin's Judicial Reform," East European Constitu
2001); Peter Solomon,
Affairs 17 (July-September
11 (Winter-Spring
tional Review
2002).
43
"What Is Shock Therapy? What Did It Do in Poland and Russia?" Post-Soviet
See Peter Murrell,
and
no. 2 (1993); Hilary
in Russia: Structural Consequences
"Voucher Privatization
Appel,
Affairs 9,
in the Second Period of Reform," Europe-Asia
Studies 49, no. 4 (1997).
Mass Response
This content downloaded from 212.175.32.130 on Mon, 26 Jan 2015 09:10:26 AM
All use subject to JSTOR Terms and Conditions
WORLD POLITICS
184
argue that the Russian leadership did itwrong and should instead have
the approaches
taken by its more
successful
But
adopted
counterparts.
this approach verges on the automatic,
the
in
rele
the
given
emphasis
vant literature on (1) leadership choice and (2) the procedure of identi
cases
the optimal
fying
approaches
by comparing
of success. Both arguments
in
differences
ignore
contexts
not
those different
different
produce
just
to the same
also different
attached
consequences
with
variable
context
records
and the ways
of choices but
menus
choices.
can be
as can
Ignoring
highly misleading,
leaving the notion
too vague. Here
to shift from the ab
of context
iswhere
it is necessary
stract to the concrete:
Poland
and Russia. The Polish
by comparing
context
that came to power in late summer 1989 had the rare
leadership
luxury
of being able to make
that could be at once easy, yet radical. I
choices
use both
in Poland had created over
adjectives
advisedly. Communism
a
consensus
time
liberal politics
and economics.
popular
supporting
consensus
This
had multiple
the
origins?for
example,
long history of
Polish nationalism, which had been shaped in part by foreign domina
tion, especially
Russian
and then Soviet;
the extraordinary
ho
national
mogeneity of Poland after the SecondWorld War; Poland's democratic
tradition (which preceded the partitions and which was further rein
forced
domination
through subsequent
of
the
communist
vulnerability
regime
pendence on the Soviet Union
polity and the economy
When
the communist
by authoritarian
in Poland, given
states); and the
its political
de
and its failure to fully Stalinize the
and to constrain
system made
popular
its formal
unrest.
departure
in the sum
mer of 1989, the newly elected Polish leadership was in the distinctive
position of being both liberal in its outlook and liberated from con
straints.
roundtable
It was
not
accords
just that Soviet power had
a
defined
transition
two-stage
retreated
to
or that
democracy.
the
It was
also that the newly elected Polish leadership enjoyed a largemandate,
outcome
in part to the
of the June 1989 semicom
unexpected
elections
and
the
and
support of the communists
petitive
subsequent
in
Gorbachev
for the formation
of a Solidarity-led
government
August
of the same year. At the same time, with
reform,
respect to economic
soon be up,
rent seekers,
Polish
the
that
would
game
long suspecting
thanks
as
as the mid-1980s
to
to reap
had begun
themselves
reposition
early
benefits
from the more
economic
liberalized
order to come. They were,
as
in short, better understood
than as antagonists.
partners
was
to
In this way, an unusually
in place in Poland
large coalition
a transition
eco
to
in
with
support
conjunction
sweeping
democracy
crisis of the
nomic
reforms
also aided by the long-term
(which were
This content downloaded from 212.175.32.130 on Mon, 26 Jan 2015 09:10:26 AM
All use subject to JSTOR Terms and Conditions
RETHINKING
Polish
economy
and the failure
a
185
DEMOCRATIZATION
of earlier
rounds
to both
in the face of
of reform
and the
stalemate).
continuing
political
was
Polish
nationalism?which
transition,
however,
political
length
Lest there be any doubt about the enormous
ened popular horizons.44
state socialism,
Poland's
radical break with
political
capital supporting
note that the finance minister,
Leszek Balcerowicz,
faced no opposition
to his economic
he consulted with other mem
reform package when
Critical
bers of the government
did not
publics
and prolonged
rebel when
the economic
and with deputies in the Sejm; that Polish
the reforms
economic
an
unexpectedly
sharp
produced
to democ
and that the transition
downturn;
more
than the roundtable
accords had pro
racy proceeded
quickly
a rare
in the
Polish
had
whether
leaders
Thus,
opportunity,
jected.
in par
in general or of new democracies
annals of democratic
politics
a radi
considerable
ticular, to combine
power and public support with
cal reform agenda.
In Russia, by contrast and for many
the weakness
reasons, including
of
the
of Russian
national
and
the
collapse of the So
identity
dynamics
and among
less consensus
viet state, there was much
among publics
elites
about
movement
either
toward
the regime or the state-in-formation.
the new order was compromised
a result, the
of
and, because
rent
of
seeking.
As
solicitous
and economically
that, necessarily
politically
was
sta
on a train
knew
Rather
than jumping
leaving the
they already
to
see
were
as
rent
in
in
Russia
seekers
tion,
Poland,
waiting
happened
In
it was heading.
whether
there even was a train and, if so, where
to recall the earlier discussion,
from the uncer
they benefited
doing so,
economic
its
transition
and
of
the
Russian
consequences?partial
tainty
reform and partial democracy.45
version
the Russian
Indeed,
or even with Gorbachev.
Yeltsin
did not
of this process
It was also characteristic
nev era, where
central control over politics
in order to prolong
the regime and keep
rather than pushing
them outside, where
originate with
of the Brezh
and economics
rent
was
seekers within
could make
reduced
the fold
even more
they
Brezhnev
gave up power in order to keep it?and
Simply put,
the lack
of the regime, the state, and the bloc, along with
the weakness
the Brezhnev
of the economy
luster performance
era, all testi
during
were. Indeed, one reason
fied to how costly those compromises
why the
so
economic
Russian
and
Soviet bloc disintegrated
performance
quickly
mischief.
44
onNation, Econ
See Rawi Abdelal, Economic Nationalism
after Empire: A Comparative Perspective
Press, 2001).
omy and Security in Post-Soviet Eurasia (Ithaca, N.Y: Cornell University
45
"Winners Take All:
The phrasing, as well as some of the logic, is borrowed from Joel S. Hellman,
in Postcommunist
The Politics of Partial Reform
Transitions," World Politics 50 (January 1998).
This content downloaded from 212.175.32.130 on Mon, 26 Jan 2015 09:10:26 AM
All use subject to JSTOR Terms and Conditions
186
WORLD
has been
taining
POLITICS
so
poor for the past decade
the regime,
the state,
power,
was
Brezhnev's
and
the bloc
legacy
of main
all
by undermining
four.46
or an elite consensus,
then, the
popular
as
"choice" in Russia, more
the
Brezhnev
in
was,
era,
recently
during
contrast
to Poland,
neither
stark nor simple. Instead,
choices were de
reverse Balcerowicz's
fined by "ordinary
apt characteriza
politics"?to
tion of the "extraordinary
that shaped the early stages of the
politics"
In the absence
Polish
transition
Daniel
Treisman
of either
a
to
and capitalism.47 As Andrei
Shleifer and
democracy
to
have argued with
economic
reform, choice
respect
in Russia was severely circumscribed.
As they generalized:
"The task of
a reformer
state is to
to
more
in a weak
stakeholders
give up
persuade
rents
in
inefficient
of
for
less
ways
socially
receiving
exchange
socially
same holds true in the
task
costly payoffs."48 The
political
sphere. The
state is to
in a precarious
of a democratizer
and
stake
regime
persuade
economic
lobbies, and the
holders?say,
regional governors,
powerful
some of their
in
communists?to
for a
give up
political power
exchange
more
stable order.
The
contrasts
between
Russia
and Poland
suggest
a more
general
on elite
emerge
problems
scholarly
emphasis
so central to the
recent democratization
(and economic
study of
the recent preferences
of the discipline
reform) and so in keeping with
to search for
and
and to privilege more
immediate
strategies
optimal
over those that are more
obvious
influences
and
removed
subtle. One
point.
choice,
Two
from
is a failure
the
as choices
decisions
among
a
es
is
competing
opportunity
tendency,
pronounced
as a
is
when
"unusual
available
standard,
comparative
pecially
politics"
to overestimate
choice
what
leaders can do. The first oversimplifies
problem
to
recognize
costs. The
most
other
the thicket of constraints. The
second
ignoring
into heroes or villains?a
sion makers
characterization
while
the excessively
democratization.
Such
impact
voluntaristic
a focus has been
of leaders during
focus
of much
transforms
that
of the literature
deci
to
speaks
on recent
on the basis of both the
justified
extraordinary
the notion of uncer
transition
the
period and
46
of the Brezhnev Era: The Rise and Fall of Corporatism,"
Valerie Bunce, "The Political Economy
British Journal of Political Science 13 (January 1983); idem, "The Empire Strikes Back: The Evolution
39 (Winter
of the Eastern Bloc from a Soviet Asset to a Soviet Liability," International Organization
1984-85).
47
See Bakerowicz,
Socialism, Capitalism, Transformation
(Budapest: Central European University
Press, 1995).
48
Shleifer and Treisman, Without aMap: Political Tactics and Economic Reform in Russia.
(Cam
MIT Press, 2000), 4.
bridge:
This content downloaded from 212.175.32.130 on Mon, 26 Jan 2015 09:10:26 AM
All use subject to JSTOR Terms and Conditions
RETHINKING
187
DEMOCRATIZATION
as a reduction
on
of constraints
has been interpreted
tainty?which
as noted earlier in this article, the
of po
leaders. However,
privileging
are both ques
of high uncertainty
litical leaders and the presumption
more room
while uncertainty
tionable. Moreover,
logically gives leaders
and eco
it also curtails
their ability to shape political
for maneuver,
the same reason?insti
nomic developments.
Both hold for precisely
tutional fluidity during transitional periods.49
If elite
tween
choice
needs
the consolidation
so does the
to be reconsidered,
be
relationship
reason is
of democracy
and its survival. One
of a regime and the quality of a regime
not be subsumed
that should
categories
that "the stability
cally separate
word?consolidation."50
The
other
reason
can be deduced
sustainability
in democratic
of
democracy
consolidation.
could
At
that is
a for
sur
of the uncertainty
to reverse
is the case, then we might want
between
consolidation
tion regarding
the relationship
formula
the usual
be
and
sustainabil
as
an
it could
be
construed
the same
the
from
earlier:
If that
ity. The
investment
analyti
one
under
it is a democracy
minus
the compromises,
of Russian
democracy,
presented
Its flaws reflect
but not fleeting.
as a consequence
mal pact, that were
required
this
transition.
rounding
ledger
flawed
are
time,
in
Russia
may explain
democracy
argued
it
is
the demo
has
survived.
Russian
Thus,
precisely
democracy
why
ex
this democratic
in Russia
that may have prolonged
cratic deficits
can
most
be
This
readily
by
supported
periment.
interpretation
to Yeltsin's
have
that
the
alternatives
might
policies
only
considering
a
furthered democracy?a
capital
capable, stable, and well-functioning
a
if Yeltsin
had proceeded
ist economy. What
party,
by, say, building
over
state control
the
democratic
institutions,
expanding
strengthening
a
course on economic
for
reforms?
the
and
Assuming
staying
regions,
moment
have been carried out, the
that these actions could somehow
rent seekers. What
would
have been to alienate
logical consequence
was a coalition of authoritarians,
economic
ensued
have
powerful
might
are
di
that
and
decentralizers?three
lobbies,
currently
key groups
a
walkout?on
in turn, could have produced
vided, not united.51 This,
the state, and democracy.
the economy,
that the failure
to consolidate
49
in Hun
Postcommunism
and the Transition:
Valerie Bunce and Maria Csanadi,
"Uncertainty
and Societies 7 (Spring 1993).
gary," East European Politics
50McFaul (fn. 31), 310.
51
in
see
of Authoritarianism
On the instability of these coalitions,
Philip Roeder, "The Rejection
both in Anderson
the Soviet Successor States," and Hanson,
Consolidation,"
"Defining Democratic
et al. (fn. 9).
This content downloaded from 212.175.32.130 on Mon, 26 Jan 2015 09:10:26 AM
All use subject to JSTOR Terms and Conditions
188
WORLD
Yeltsins
were
therefore,
options,
POLITICS
threefold:
to resist
a liberal
revolu
tion (like the Polish political stalemate of the 1980s), to embrace this
revolution fully (as in the Czech Republic, Slovenia, or the Baltic
fashion.
It can be argued on
states), or to proceed, but in compromised
context
the basis of the Russian
that the third approach was best posi
because
of that, sustainable
tioned to yield a flawed but, perhaps
de
was
but
Thus,
durable;
mocracy.
democracy
shortchanged
capitalism
was
to
distorted
but
and the state, while
implanted;
nonetheless
survived.
challenges,
weak
and vulnerable
Conclusions
on democratization,
Research
mance
litical
the founding
and perfor
particularly
a
is
the
literature about
choices po
democracies,
largely
It is
and the consequences
of those choices.
have made
of new
leaders
also
largely
and
America
a literature
southern
based
on
The
the
return
to democracy
of this article
in Latin
has been
purpose
Europe.
to
the
focus on elites and the general
and
question
complicate
to
demo
about transitions
izations
that have been made
democracy,
so
I
done
cratic consolidation,
and democratic
have
sustainability.
by
an additional
to the
twenty
region
adding
empirical
equation?the
both
seven
to
countries
Several
up the Eurasian
postcommunist
region.
seem to
to
transitions
democracy
emerged. First,
to
the
the
with
respect
uncertainty
surrounding
that make
conclusions
vary
considerably
variance
process. This
in turn affects
of transition
and
the strategies
was
In the postcommunist
it
assumed
that
their payoffs.
region
widely
was
these transitions
the uncertainty
unusually
high,
surrounding
cases of a democratic
to
most
in
for
the
absence
past
given,
example,
of state so
economic
and political penetration
gether, the extraordinary
state
the
tensions
and
democratization,
cialism,
among
seeming
a
and
the
of
the construction
economy,
radically
capitalist
building,
of the state to the international
system. It turns
changed
relationship
a
in
the
countries
number
of
that
for
out, however,
region the transition
was
to
the
for two reasons. First,
in fact not so uncertain,
democracy
was
was eliminated
there
from the transition.
Second,
present
military
a
that gained
strength from popular mobilization
opposition
powerful
the state) (as with
the Baltic,
the
also
(often
against
against
regime
who
communists
reform
and Polish
Slovenian,
Czech,
cases) and/or
to
the
committed
(as with
democracy
opposition
and
Poland,
Slovenia,
countries,
Hungary).
was
in all of
the transition
Because
lower, moreover,
uncertainty
collaborated
with
an
Baltic
This content downloaded from 212.175.32.130 on Mon, 26 Jan 2015 09:10:26 AM
All use subject to JSTOR Terms and Conditions
RETHINKING
cases
these
DEMOCRATIZATION
189
a
state socialist
ex
sharp break with the
past?for
a
elections
that
the
gave
ample,
through
founding
opposition
large
in constructing
democratic
mandate,
institutions,
rapid progress
quick
introduction
of far-reaching
economic
of the
and, in most
reforms,
produced
of a new
cases, the construction
more uncertain where
mass
the military
on
focused
mobilization
state.
was
leaving
the communists
transition was far
contrast,
in
the
transition, where
engaged
By
state but not
de
building
were able to command
consid
the
and/or where
mocracy,
erable
in the first election.
As
support
was
authoritarian
less
definitive?in
past
a result,
the break with
the
terms of both
lead
political
ership and public policy.
contrasts
These
most
successful
have
is that, while
One
the
implications.
most
in the South
the
involved
bridging,
in the East
involved breakage.
it is pre
Indeed,
transitions
successful
several
transitions
cisely the bridging approach in the East that produced the most fragile
democracies.
The
breakage?and
sure reflected
other
the costs
differences
is that
the
and benefits
contrast
between
bridging
of each approach?in
large
and
mea
in uncertainty.
is that mass mobilization
can
a very
conclusion
play
posi
as it did, for
in the transition,
in the Baltic, Polish,
example,
cases and, most
and Slovenian
in Serbia and Mon
Czech,
recently,
mass
can reduce uncer
This
is
because
mobilization
tenegro.
largely
as well
the
of
the
communists,
tainty, thereby
influencing
preferences
as the division
of power between
them and the opposition.
Another
tive role
Nationalist
influence
mobilization
the democratic
and the disintegration
of the state can also
occurs seems to reflect a
Whether
this
project.
such protests first arose when
whether
the regime and
key distinction:
state were
or whether
at that time were
the
demonstrations
unraveling
the culmination
of a longer history
In the first case,
of such protests.
which
describes
Baltic
the
and
Slovenia,
countries,
Russia, Ukraine,
the transition
sustainable
albeit
democratic
Moldova,
orders,
produced
of varying
in every transition where
nationalist
quality. By contrast,
a
new
had
both
the
old
and the
state, as well as
protest
longer lineage,
the democratic
contestation.
continuing
project, experienced
This
into two
nationalist
leads to another
groups?where
mobilization
conclusion.
If we
the transition
and where
was
itwas
divide
the twenty-seven
accompanied
not?we
find
cases
by significant
two sets of sto
ries. As already noted,
the first story is about the timing of nationalist
In the second group, the key issue is the
mobilization.
strength of the
outcome
is
the
which
for
of the first
indicated,
opposition,
example, by
election.
This content downloaded from 212.175.32.130 on Mon, 26 Jan 2015 09:10:26 AM
All use subject to JSTOR Terms and Conditions
190
WORLD
This
brings
the amendments
POLITICS
us to our final
set of arguments.
It is true, when
that political
leaders?their
discussed,
already
their power, and their actions?are
ences,
the sustainability
of democracy.
in different
countries
leaders
that similar
choose;
same consequences;
generally
the state
choices
However,
have
the
critical
it does
not
then
same menu
follow
that
from which
to
have
the
contexts
in different
adding
prefer
to the
and
founding
necessarily
choices
that are
are, as a result, optimal
or that
the democratic
compromising
project and
that there
applicable;
and after the transition
reduces the sustain
necessarily
are two basic distinctions
of
There
here: between
ability
democracy.
conditions
and strategies
and between
the consolidation
of democracy
during
and sustaining
In the first
more
it.
the key point
distinction,
more uncertain?than
constrained?or
is that
some
transitions
are
and
it is
others,
precisely
to
the degree of uncertainty
that defines both the strategies
available
the
leaders
and
of
those
consequences
political
strategies. Thus,
"easy"
feature very different matrices
of choices
transitions
and payoffs
than
were
In the postcommunist
do "hard" transitions.
world
transi
there
was
the opposition
and the au
tions, as noted above, where
powerful
or collaborative
as a result,
thoritarians
and where,
either marginalized
there could be a radical break with
the past. These
then pro
transitions
duced a quick consolidation
of both democracy
and capitalism
and,
even the state.
state
set
when
also
accompanied
by
disintegration,
They
the standard
for what
well as in politics.
Most
transitions
however.
category,
was a
compromised
did not necessarily
constituted
the ideal approach?in
economics,
as
in the postcommunist
did not fall into this
world
was
Instead, uncertainty
higher, and the best result
this
and state. Nonetheless,
democracy,
capitalism,
mean
that
leaders
contexts
the
adopted
the "wrong" conditions.
over
and opted for breaking
in these
faced
Rather,
strategies.
they merely
if they had pretended
otherwise
more
successful
their
counterparts,
bridging,
thereby emulating
they
no
or
with
well
have
ended
state,
very
up
democracy,
might
capitalism.
are defined
not
Thus,
strategies and their particular payoffs
by contexts,
wrong
Moreover,
those cases have similar contexts.
by other cases?unless
the consolidation
between
This
and
leads, in turn, to the relationship
true that consolidated
It
is
de
of
democracy.
certainly
sustainability
not
are
to
it
follow
sustain
themselves.
But
does
mocracies
very likely
policies
necessarily
or that
are
less sustainable
necessarily
of democracy
the consolidation
that compromise
case
as the Russian
from its sustainability.
Indeed,
democracies
that unconsolidated
and behaviors
detract
This content downloaded from 212.175.32.130 on Mon, 26 Jan 2015 09:10:26 AM
All use subject to JSTOR Terms and Conditions
RETHINKING
191
DEMOCRATIZATION
it may be
the limits to democracy,
suggests,
precisely
state and
and the policies
that contributed
capitalism,
that sustain all three.
as well
as to the
to those
limits
are now
to address some issues of broader concern.
in a position
the cases chosen do indeed seem to de
First, as this article reinforces,
termine
is particularly
drawn. This
the case, one can
the conclusions
We
case
argue, when
selection
case selection
reduces
also seems
variation
in
dependent
variables.
to
and therefore
shape assumptions
as
it can be
in
the
noted
Second,
introduction,
analytical
approaches.
we have discovered,
to restrict our
As
reach.
costly
regional
expanding
can introduce
new variables
new issues, while
horizons
and
regional
common
and arguments.
How
challenging
approaches,
assumptions,
Moreover,
the South and the East that
the repeated contrasts between
in this article, an obvious
itself: do these
presents
question
emerged
are
contrasts mean
defined?
that political
regionally
dynamics
to concur with
this statement. After
It is tempting
all, for the post
cases in
is logical?
notion
communist
the
of
effects
regional
particular,
the structural
similarities
for instance,
forged by the political
given,
ever, given
and the Soviet bloc; the common
of state socialism
economy
origins of
states
of the ethnofederal
all the new states as a result of disintegration
in
in the region along
the timing as
lines; and the similarities
republican
to
in the transition
How
well as the key players
involved
democracy.
notion
I
the
that
would
nonetheless
ever,
argue against
dy
political
to
namics respond
regional effects.
First
a
obviously,
major
is their extraordinary
it is ironic that the variable
and most
communist
cases
for analyzing
the post
not their similarities.
variability,
rationale
in
of authoritarianism
practices
and southern Europe
and the variable
timing of their
seem to have
to
in tran
less
variation
transitions
democracy
produced
of the democratic
sition dynamics
and in the quality and sustainability
we see in the
context.
like urban
Second,
project than
postcommunist
and even some nongeographical
such as
rural distinctions
cleavages,
that
and
class,
gender
analysts habitually
employ, region only begs the
Indeed,
Latin America
at work. As Adam
Przeworski
and
actually
com
more
the
of
than
Teune
purpose
thirty years ago:
Henry
argued
with variables.52
is to replace place-names
Indeed, this
parative analysis
to do
the discus
is precisely what
this article has attempted
by framing
sion in terms of variations
mobilization,
(1) in the timing of nationalist
question
about
the factors
(2) in the historical role of themilitary in politics, (3) in the strength of
52
Przeworski
and Teune,
The Logic of Social Inquiry
(New York: John Wiley,
1970).
This content downloaded from 212.175.32.130 on Mon, 26 Jan 2015 09:10:26 AM
All use subject to JSTOR Terms and Conditions
192
WORLD
POLITICS
the opposition, (4) in the uncertainty built into the transition, and (5)
in the range of policy options available to political leaders and their
payoffs.
Thus,
a summary of factors that have taken on geo
region ismerely
form. For this reason and because regions can provide not just
graphical
new factors and variation
in those factors, cross-regional
studies can be
or
in
those
and ar
quite helpful
contesting
complicating
assumptions
that were
from the analysis of one or several similar
derived
are very different
This
is
the case when
regions
particularly
regions.53
in culture, historical
from one another
and relationship
development,
new
to the international
causal considerations
system; when
they add
guments
to the
of
dynamics
analysis; when
they vary the timing of the political
in
variation
and
when
evidence
considerable
interest;
they
dependent
not because
It is precisely
variables.
for these reasons?and
region itself
our
to
it is advisable where possible
matters?that
expand
geographical
its
This
is particularly
the case for democratization,
horizons.
given
reach.
global
53
See, for example, Doug McAdam,
Sidney Tarrow,
Press, 2001).
(Cambridge: Cambridge University
and Charles
Tilly, Dynamics
This content downloaded from 212.175.32.130 on Mon, 26 Jan 2015 09:10:26 AM
All use subject to JSTOR Terms and Conditions
of Contention
Presidential Popularity in a Hybrid Regime:
Russia under Yeltsin and Putin
Daniel Treisman
University of California, Los Angeles
In liberal democracies, the approval ratings of political leaders have been shown to track citizens’ perceptions of the state
of the economy. By contrast, in illiberal democracies and competitive autocracies, leaders are often thought to boost their
popularity by exploiting nationalism, exaggerating external threats, and manipulating the media. Using time-series data, I
examine the determinants of presidential approval in Russia since 1991, a period in which leaders’ ratings swung between
extremes. I find that Yeltsin’s and Putin’s ratings were, in fact, closely linked to public perceptions of economic performance,
which, in turn, reflected objective economic indicators. Although media manipulation, wars, terrorist attacks, and other
events also mattered, Putin’s unprecedented popularity and the decline in Yeltsin’s are well explained by the contrasting
economic circumstances over which each presided.
I
n liberal democracies, the approval ratings of political leaders play an important role in the interaction between citizens and their governments.1 Like
elections—but far more frequently—they communicate
to officials what the public thinks of their performance.
Moreover, the messages such polls send have been shown
to make sense. In various Western democracies, leaders’
ratings tend to rise and fall with the country’s economic
performance. Respondents appear to hold their leaders
responsible and reward them for effective economic management with approval, much as, when voting retrospectively, they repay competent and honest incumbents with
reelection.
But what role do such ratings play in illiberal democracies and competitive autocracies, where formally democratic institutions coexist with authoritarian elements? In
such states, public opinion is often viewed as more a product of political manipulation than an input into politics.
Even when the polls themselves are trustworthy, they are
thought likely to capture the effects of rulers’ political theater rather than reasoned evaluations rooted in material
reality. Citizens, fed distorted information by an unfree
press and cynical about the possibilities for participation,
are expected to focus on image and personalities rather
than to soberly evaluate performance by studying economic statistics (Ekman 2009). To rekindle their appeal
periodically, incumbents invoke cruder forms of nationalism, exaggerate external threats or terrorist dangers, or
even start military engagements (Mansfield and Snyder
1995).
One country where these issues arise particularly
starkly is Russia. Since the first competitive presidential
election there in 1991, presidents’ ratings have careened
wildly. Russia’s first president, Boris Yeltsin, started out
extremely popular; in September 1991, 81% of citizens
approved of his performance. When he left office eight
years later, the proportion had dropped to 8%. His successor, Vladimir Putin, saw his approval shoot up from
31% in August 1999 to 84% in January 2000. In his eight
years as president, his rating rose as high as 87% and never
fell below 60%.
To some, these extreme swings reveal the country’s
political immaturity. Journalists attributed both wars in
Chechnya to the desire to boost a Kremlin-backed candidate’s electoral prospects. Putin’s appeal has been explained as the result of an artificially cultivated image
as a “tough” leader, the trauma of terrorist bombings
in late 1999, and the Kremlin’s enhanced control and
Daniel Treisman is Professor of Political Science, University of California, Los Angeles, 4289 Bunche Hall, Los Angeles, CA 90095-1472
([email protected]).
I thank Yevgenia Albats, Robert Erikson, Lev Freinkman, Tim Frye, Scott Gehlbach, Vladimir Gimpelson, Patrik Guggenberger, Sergei
Guriev, Arnold Harberger, John Huber, Rostislav Kapelyushnikov, Matthew Lebo, Grigore Pop-Eleches, Brian Richter, Richard Rose, Jim
Snyder, Konstantin Sonin, Stephen Thompson, Jeff Timmons, Josh Tucker, Andrei Yakovlev, Alexei Yudin, Katya Zhuravskaya, and other
participants in seminars at the Higher School of Economics, Moscow, and Columbia University, as well as the editor and three anonymous
referees for comments or helpful suggestions, and the UCLA College of Letters and Science for research support.
1
The dataset, statistical analysis, and a web appendix for this article are available on the author’s website at www.sscnet
.ucla.edu/polisci/faculty/treisman/.
American Journal of Political Science, Vol. 55, No. 3, July 2011, Pp. 590–609
C 2011,
590
Midwest Political Science Association
DOI: 10.1111/j.1540-5907.2010.00500.x
591
PRESIDENTIAL POPULARITY IN RUSSIA
manipulation of the media. Such interpretations seem to
fit the fluidity of the political scene, where parties come
and go and politics has often been seen as primarily the
clash of personalities.
But there is another possibility. The volatility of presidents’ ratings and vote shares might merely mirror volatility in the economy. Russians might be responding to
perceived conditions as logically as do their Western counterparts. Swings in incumbents’ approval might reflect not
Kremlin manipulation or the capriciousness of voters but
their admittedly crude attempts, in an environment of
uncertainty and unresponsive government, to hold their
leaders to account.
Although there are many conjectures about what
causes change over time in the popularity of Russian presidents, few have attempted to confront these systematically
with evidence.2 I use statistical techniques and time-series
survey data to do so. I look first at what polls reveal directly
about the determinants of presidential popularity. Then,
with error correction regression models, I test which factors help predict the trajectories of presidents’ ratings.
Understanding what affects the popularity of Russian presidents is important for several reasons. Not least,
it matters for Russian politics. If building support for
an incumbent requires only projecting a certain image,
it makes sense for Kremlin operatives to monopolize the
mass media. If fighting Chechen terrorists boosted Putin’s
appeal, such threats are likely to be dramatized as elections
approach. However, if the key element was economic recovery, implications are more benign: leaders will need to
master the art of economic management.
Although Russia is unique in certain ways, the results suggest hypotheses and modes of analysis relevant
to other hybrid regimes. Debates continue over the influence of economics and charismatic populism over voting
and presidential approval in Latin America (Dominguez
and McCann 1996; Weyland 2003). Does Hugo Chavez’s
appeal in Venezuela rest on ideological support for his
“Bolivarian Revolution” or on oil-fueled prosperity? Did
Fujimori’s support in Peru owe more to economic conditions or his counterinsurgency efforts (Arce 2003)? Scholars have also pondered what shapes presidential popularity in the young democracies of Asia, exploring, for
instance, whether the dips in Indonesian President Yudhono’s rating in 2008 were caused by falling oil prices or
the image of the president as weak and indecisive (Mietzner 2009, 147).
To preview the results, I find that Russians resemble
voters in developed democracies in how they judge their
leaders. Their evaluations closely follow economic con2
One exception is Mishler and Willerton (2003).
ditions. Perceptions of the state of the Russian economy
and of families’ own finances do a good job of predicting both the decline in Yeltsin’s rating, and the surge
and plateau in Putin’s. Although leaders have tried to
manipulate perceptions, the statistical evidence suggests
such efforts have had limited effects. At times—notably
during certain election campaigns—Russians’ views of
the economy were rosier than warranted. But in general,
perceptions were well predicted by objective economic
indicators.
In this article, I study what determines change in
the average popularity of Russian presidents over time.
Other papers, using cross-sectional survey data, have explored what types of individuals supported Yeltsin or
Putin at given moments. Such studies, like this one, tend
to find that economic factors were important (Colton
and Hale 2008; Duch 1995; Hesli and Bashkirova 2001;
Miller, Reisinger, and Hesli 1996; Rose 2007b; Rose, Mishler, and Munro 2004; White and McAllister 2008). While
this might seem reassuring, it is important to remember
how cross-sectional and time-series analyses differ. They
do not, as sometimes thought, generate potentially competing evidence on the same question. Rather, they address different questions. Cross-sectional analyses show
what categories of citizens favored the incumbent at a
particular time. They reveal little about changes in aggregate support. For that, we need time-series or panel data.
Of course, time-series regressions are unable to assess the
importance of characteristics that vary across individuals
but not over time.3
The only time-series study of Russian presidential
approval of which I am aware is Mishler and Willerton
(2003), which examined data up to 2000, and so was not
able to draw strong comparisons between the Yeltsin and
Putin periods. I extend their analysis, and, using the more
extensive data now available, draw somewhat different
conclusions.
Data on Presidential Approval
The main data I analyze are from a regular, face-to-face
survey conducted by the Russian Center for Public Opinion Research (VCIOM) until 2003, and then by its successor, the Levada Center. VCIOM, founded in 1988, was
directed from 1992 by Yuri Levada, a sociologist who
had been fired from Moscow State University in 1969 for
3
As Kramer demonstrated: “There is no reason whatever to expect
time-series and cross-sectional estimates of the same parameters to
be similar in magnitude; they need not even be of the same sign”
(1983, 93). For an excellent discussion of cross-sectional versus
time-series data, see Erikson, MacKuen, and Stimson (2002).
592
DANIEL TREISMAN
FIGURE 1 Presidential Approval, Russia, 1991–2008
Source: Surveys of VCIOM and Levada Center (see web appendix).
Note: Yeltsin approval is percentage of respondents saying on the whole they approve of
the performance of Boris Yeltsin. Likewise for Putin approval. Ratings on 10-point scale are
average answer to question: “What evaluation from 1 (lowest) to 10 (highest) would you give
the President of Russia (name of president)?” Missing months interpolated. Putin approval
includes his first period as prime minister.
“ideological mistakes in his lectures.” VCIOM earned a
reputation as the most professional and politically independent of Russia’s half-dozen leading polling organizations. This independence is thought to have prompted a
state takeover in 2003, in which Levada was fired. The center’s researchers set up the private Levada Center, which
continued the polls.
The surveys are of a nationally representative sample of voting-age citizens, who are interviewed in their
homes. I focus on two questions. First, the pollsters ask:
“On the whole do you approve or disapprove of the performance of [the president’s name]?” This question, similar to one on the U.S. Gallup poll, was included monthly
from late 1996.4 I use it to examine approval of President
Putin, first elected in March 2000. Since this question
was asked only occasionally before September 1996, I use
another to study Yeltsin’s popularity. This one, included
every second month from early 1994, asked: “What evaluation from 1 (lowest) to 10 (highest) would you give the
President of Russia Boris Yeltsin?”5 I analyze the average
evaluation.6 When both measures are available, the two
are highly correlated, in levels (at r = .98) and in twomonth differences (r = .91).
Figure 1 shows the data. On the left, Yeltsin bumps
his way down from a December 1990 peak to a low of
6% in early 2009. On the right, Putin glides along, consistently above 60%. No American president has equaled
Putin’s record since regular polling began in the 1930s.
Eisenhower came closest, but even his rating fell at times
into the 40s. No British prime minister has come close
since the first MORI poll in 1979.7 Nor have postwar U.S.
presidents plumbed the depths to which Yeltsin sank—the
lowest was Harry Truman’s 22% in February 1952.
A natural first question, then, is whether the results
are believable. Might Putin’s ratings have been concocted
to please the Kremlin or reflect insincere replies of intimidated respondents? There are reasons to doubt this.
6
4
Gallup asks: “Do you approve or disapprove of the way that . . . is
handling his job as President?”
I constructed a dataset including just the alternating months in
which the question was asked, from early 1994 when the regular
series began. In a previous version, I interpolated missing values;
however, given the large amount of interpolation necessary, I prefer
here to analyze the bimonthly series. I also use a bimonthly series
for the Putin period since economic explanatory variables were
available only every second month.
5
7
This question was also used by Mishler and Willerton (2003). In
the approval question, sample sizes were about 1,600; in the 10point scale questions, the size ranged from 2,100 to about 2,400.
See historical data on the U.S. Gallup poll at www.presidency
.ucsb.edu/data/popularity.php and the Ipsos-MORI polls at
http://www.ipsos-mori.com/polls/trends/ satisfac.shtml.
593
PRESIDENTIAL POPULARITY IN RUSSIA
First, it is hard to believe VCIOM was slanting results to
please Putin given its leader’s semidissident past and the
state’s takeover to punish it for insufficient loyalty. Various Western pollsters (such as World Public Opinion and
the New Russia Barometer) have worked with the Levada
group and found it highly professional. The dynamics
in VCIOM/Levada polls match those in surveys by other
organizations. For instance, in 2006–2007 the Levada 10point rating correlates at r = .93 with the share that said
in polls of the Fond Obshchestvennogo Mnenia that they
“trusted” Putin.
Second, it seems unlikely many respondents were intimidated given their critical responses on other questions. Asked in 2004 whether there was more or less corruption and abuse of power in the highest state organs
than a year before, 30% said “more,” 45% said “the same
amount,” and only 13% said “less.”8 Respondents were
not shy to give Yeltsin a 6% approval rating and Putin just
31% in 1999. Even as Russians swooned over Putin, until
mid-2007, his governments never won the approval of
more than 46%, and large majorities opposed some of his
policies, including—after the initial period—the military
occupation of Chechnya.9
Another concern is that, by limiting respondents to a
10-point scale, the surveys might be censoring the data. In
the web appendix, I show the distributions of responses
around Yeltsin’s highest and lowest points. While the distribution is reasonably symmetric for Yeltsin’s peak, respondents did cluster at the bottom in 1999. If the scale
was censoring some respondents, the effect of economic
decline under Yeltsin may actually have been greater than
that estimated here.
Possible Explanations
What might explain the path of approval in Figure 1?
In other countries, the public often rallies behind leaders at the start of wars (Mueller 1973). Both wars in the
8
Results at www.russiavotes.org, Slide 450. Twelve percent picked
“don’t know.”
9
Wyman reviews the difficulties and common criticisms of polling
in Russia and concludes that most problems are those faced by
survey researchers worldwide. He found “no evidence specific to
Russia that respondents . . . engage in self-censorship” (1997, 5–19).
To investigate this, Rose (2007a) asked respondents in 13 postcommunist countries in 2004–2005 whether they thought “people today
are afraid to say what they think to strangers.” Among Russians,
25% thought people were afraid to some extent—the lowest rate
for all 13 countries. Those who thought people afraid to talk—
and who were presumably nervous themselves—were only slightly
more likely to favor the current regime, suggesting distortions due
to self-censorship are minor.
southern republic of Chechnya began in the run-up to
presidential elections. In November 1994, Yeltsin’s aide
Oleg Lobov reportedly said that a “small, victorious war”
would “raise the president’s ratings.”10 Putin’s surge occurred as he sent troops to the republic for a second time
and promised to crush the terrorists who had bombed
four apartment buildings. “People believed that he, personally, could protect them,” Yeltsin wrote in his memoirs.
“That’s what explains his surge in popularity” (Yeltsin
2000, 338).
Surveys suggest the public viewed the two wars quite
differently. Yeltsin’s use of force in December 1994 was
widely condemned; the next month, two-thirds of respondents opposed it (Jeffries 2002, 372). By summer, 71%
disapproved of Yeltsin’s approach to Chechnya. However,
opinion soon soured on the compromise of 1996, under which terrorists regularly crossed the border to take
hostages for ransom. In October 1999, 74% favored a
major military operation against illegal armed groups in
Chechnya.11
Support, however, proved fickle. By July 2000, about
two-thirds of Russians thought Putin’s attempts to rout
the insurgents mostly or completely unsuccessful, and
the share remained above 60% until 2006. By November 2000, more Russians favored starting peace negotiations than continuing the military operation. By early
2001, fewer than one in ten respondents said they were
attracted to Putin by his Chechnya policy. If Putin’s hard
line on Chechnya helped him early on, this does not seem
to have been the case later. To capture attitudes toward
Putin’s Chechnya policy, I use a variable that measures the
percentage of respondents who favored “continuing the
military operation” in Chechnya rather than “negotiating
with the ‘fighters’.”12
Second, some saw the main reason for Putin’s popularity and Yeltsin’s dwindling appeal in their divergent
personal styles. Yeltsin—ailing, gaffe-prone, at times visibly inebriated—could hardly have seemed more different
from the disciplined, energetic, sober Putin, a former spy
and judo black belt. Polls confirm that Russians were attracted by Putin’s image of youthful vigor and put off by
Yeltsin’s physical decline. Asked what qualities attracted
them to Putin, from 30 to 47% of respondents chose
10
Lobov later denied saying this, but admitted overestimating the
odds of a quick success (Colton 2008, 290).
11
VCIOM, Omnibus poll 1995-4 and Express poll 1999-11, at
http://sofist.socpol.ru.
12
The question was not available for the first war. In a previous
version I used the percent that said “war is continuing” in Chechnya
rather than “peace is being established.” However, after correcting
a data error, preferences on Chechnya policy were more significant,
and also required less interpolation of missing values.
594
“he is an energetic, decisive, strong-willed person.” Only
9% said this of Yeltsin in January 2000. Even in 1996,
the third most frequent thing respondents disliked about
Yeltsin was that he was “a sick, weak person.”13
But did the presidents’ images explain their varying
support? Unfortunately, regular data on this were not
available; respondents were asked only occasionally about
Putin’s attractions. Governing style is also manifested in
particular incidents. One episode thought to have harmed
Putin’s image was his reaction when the Kursk nuclear
submarine sank in 2000. While the Navy brass dithered,
ignoring Western offers of help, the news showed Putin
jet-skiing on the Black Sea. Yeltsin’s frequent hospital stays
cannot have improved his image (Mishler and Willerton
2003). His drinking problem was made vivid in August
1994, when television showed him jerkily conducting a
police band in Berlin.
Putin promised from the start to restore “order” after the turbulent 1990s, to fight crime and corruption
and reimpose control over wayward local officials. A
former KGB officer, he seemed to have the appropriate
experience and connections. Could this explain his popularity? In surveys, many Russians say they favor strong
leaders and are willing to sacrifice some rights in return
for order.14 However, advocates of a “strong hand” and
other antidemocratic norms turn out to be less likely to
support Putin than to back Communist leader Gennady
Zyuganov or the ultranationalist Vladimir Zhirinovsky.
In fact, Whitefield (2005) found it was democracy supporters who favored Putin.15
Moreover, many Russians were skeptical of Putin’s
claims. Periodically, VCIOM/Levada polls asked how successful Putin had been at imposing order. On average
47% thought he had been at least partly successful; 49%
DANIEL TREISMAN
thought he had been unsuccessful.16 There was no clear
trend. On more specific questions, reports of deterioration predominated. Each year after 2000, at least 25%
more thought citizens’ personal security had worsened
than thought it had improved; at least 15% more saw decline in law enforcement than saw progress.17 Early on,
29% liked Putin because he “could impose order in the
country.” By October 2006 the share had fallen to 13%.18
Following Mishler and Willerton (2003), I created a
measure of Russians’ political mood from the question:
“Overall, how would you assess the political situation in
Russia?” I subtracted the proportion choosing “critical,
explosive,” or “tense” from that saying “calm” or “favorable.” To test whether attacking the oligarchs and exploiting nostalgia won Putin support, I created dummies
for the months after the arrest of the oligarch Mikhail
Khodorkovsky and after Putin restored the Soviet music to the national anthem. A large plurality—46% in a
VCIOM poll—favored the Soviet version.19
Respondents credited Putin with “strengthening Russia’s international position.” Between July 2000 and
March 2007, on average 65% of respondents said he
was quite or very successful at this, compared to 47%
who said this of his efforts to “introduce order in the
country.” By contrast, many thought Yeltsin had been too
accommodating toward the West, which expanded NATO
into Eastern Europe on his watch and then bombed the
Serbs over Kosovo. To capture effects of foreign affairs,
I included dummies for the Kosovo bombing, the 9/11
attack, which prompted a pro-American turn on Putin’s
part, and the 2003 U.S. invasion of Iraq. The Kosovo and
Iraq wars were both unpopular in Russia and might have
sapped support for presidents viewed as too cozy with
Washington. Under Putin, I also included the share that
16
13
VCIOM Express 1996-3, 15-20 February, 1996, 1,584 respondents, at http://sofist.socpol.ru.
14
Does this reveal a cultural predilection for authoritarian rule or
just a response to extreme conditions? After 9/11, large majorities
of U.S. and British respondents were ready to trade civil liberties
for security. In a YouGov poll of British adults, 70% said they were
“willing to see some reduction in our civil liberties in order to
improve security in this country” (“Observer Terrorism Poll: Full
Results,” Observer, September 23, 2001). In a New York Times/CBS
poll, 64% of U.S. respondents agreed that in wartime “it was a good
idea for the president to have authority to change rights usually
guaranteed by the Constitution” (Robin Toner and Janet Elder,
“A Nation Challenged: Attitudes; Public Is Wary but Supportive
on Rights Curbs,” New York Times, December 12, 2001). See Hale
(2009) for a debunking of the common claim that Russians are
undemocratic.
15
Other polls confirm that Putin is not more popular among supporters of authoritarian government (Morin and Samaranayake
2006; Rose, Mishler, and Munro 2004).
See “Prezident: otsenki deatelnosti,” Levada Center, at
http://www.levada.ru/ocenki.html.
17
Levada Center press release, at http://www.levada.ru/press/
2007120703.html.
18
I used a measure of the share of respondents who said Putin had
been very or quite successful in creating order in Russia minus the
share who thought he had been mostly or completely unsuccessful.
As this question was asked only 14 times, this required shortening
the data period and interpolating two-thirds of the data. Similar
problems apply to the international affairs variable. Given the need
for extensive interpolation, I include these variables with strong
reservations; however, some previous readers thought it important
to do so.
19
VCIOM Express poll, 2000–22, 27–30 October, 2000, 1,600 respondents, and Romir Omnibus poll, 2000–11, 1–30 November,
2000, 2,000 respondents. Results available at http://sofist.socpol.ru.
In a previous version, in which I interpolated more data, I was able
to examine the effect of the constitutional crises of April and October 1993. Avoiding such interpolation, available data now begin
in 1994.
595
PRESIDENTIAL POPULARITY IN RUSSIA
said he was strengthening Russia’s international position
minus the share that thought him mostly or completely
unsuccessful in this.
To his opponents, Putin’s ratings simply showed
the Kremlin’s increasing dominance of the press. In
the words of Garry Kasparov: “You cannot talk about
polls and popularity when all of the media are under
state control” (Remnick 2007). Under Putin, state companies or loyal businessmen took over the last independent national television networks. Strong criticism
of the president—although not of the government—
disappeared from broadcasts. Based on a 1999 survey,
White, Oates, and McAllister (2005, 192) concluded that
media bias helped secure Putin’s victory: “The decisive
factor in this dramatic reversal of fortunes appeared to
be the media, particularly state television.” Measuring
change in press freedom was difficult. Lacking a more sophisticated gauge, I used Freedom House’s index of press
freedom, with the annual value used for each month of
the relevant year. I also created a variable for the month
of the state takeover of NTV, the network previously most
critical of the Kremlin.
Finally, economic performance—and, especially, perceptions of it—has been shown to affect approval of incumbents in the United States, France, and Britain (Clarke
and Stewart 1995; Erikson, MacKuen, and Stimson 2002;
Lafay 1991; Sanders 2000). In Russia, scholars have found
economic influences on voting (Colton 2000; Tucker
2006), and these might also shape presidential popularity. Such effects might be retrospective or prospective and
might focus on respondents’ own circumstances or their
views of conditions nationwide. Previous studies of postSoviet countries found evidence of all four types of economic influences. Hesli and Bashkirova (2001), looking at
cross-sectional surveys, found all four effects affected support for Yeltsin. Mishler and Willerton (2003), focusing
on time-series data, noted the influence of retrospective
evaluations of the national economy and family finances.
Rose, Mishler, and Munro (2004, 209) found in a crosssection that views of current economic performance were
the strongest predictor of support for the political regime.
Data were available in the VCIOM/Levada surveys to
construct three variables. For retrospective evaluations of
the national economy, I use the question: “How would
you assess Russia’s present economic situation?” For retrospective assessments of personal finances, I use: “How
would you assess the current material situation of your
family?” For each of these, I subtracted the shares saying “very bad” or “bad” from those saying “very good”
or “good,” ignoring those who said “in between” or
“don’t know.” For prospective evaluations of the national
economy, I use: “What do you think awaits Russia in the
economy in the coming several months?” I subtracted
the percentage anticipating decline from that expecting improvement.20 Unfortunately, no question captured
prospective evaluations of personal finances for a comparable period. I use a dummy for the shock of the August
1998 financial crisis and another for January 2005, when
Putin introduced a reform to replace in-kind benefits
such as free bus tickets and drugs for pensioners by cash
grants. This provoked major demonstrations of benefit
recipients, who complained that the compensation was
insufficient.
Explaining Presidential Approval:
Analysis
Before analyzing the presidential ratings, some statistical issues must be addressed. As is well known, OLS regressions on nonstationary data may produce spurious
results. I therefore examined whether the approval data
and the time-series explanatory variables were stationary,
i.e., I(0); had a unit root, i.e., I(1); or were something
in between, i.e., fractionally integrated, I(d) where 0 <
d < 1.21 Table 1 shows test statistics for these series. I
use the augmented Dickey-Fuller (ADF) and the PhillipsPerron tests to test the null hypothesis of a unit root,
and the Kwiatkowski et al. (KPSS) and Harris-McCabeLeybourne (HML) tests to test the null of stationarity. I
treat the two presidencies separately because, examining
Figure 1, it seems likely the underlying process changed
between their tenures.22
The tests have weak power and do not always agree.
In almost all cases, the HML test—but not the KPSS
test—suggests the series are not stationary.23 In almost
all, the ADF and Phillips-Perron tests cannot exclude the
possibility of a unit root.24 Given the likelihood that most
20
These variables were available in a continuous bimonthly series
from early 1994, and irregularly before then.
21
On the analysis of fractionally integrated time series, see, for
instance, Box-Steffensmeier and Smith (1996).
22
And analysis using the 10-point scale for both periods (not shown
here) confirms that the coefficients on key variables changed between the two presidencies. Economic perceptions, although still
highly significant, had somewhat smaller coefficients under Putin.
It is intuitive to suppose that Russians acclimated psychologically to
the stable growth under Putin and were less sensitive to short-run
changes than they were to the large, sustained, irregular economic
declines under Yeltsin.
23
The exception is expectations in the Putin period, for which the
HML test cannot reject stationarity.
24
The exceptions are economic expectations under Putin, for which
the Phillips-Perron, but not the ADF test, rejects the null of I(1),
596
DANIEL TREISMAN
TABLE 1 Testing for Stationarity and Cointegration
A. Under Yeltsin (May 1994–December 1999, Bimonthly Data)
ADF test of I(1)
Phillips-Perron test
of I(1)
KPSS test of I(0)
HML test of I(0)
Estimate of d
Estimate of d for
residuals of
regression of
Yeltsin rating on
this variable
Yeltsin
Rating
Current
Economy
Family
Finances
Russia’s Ec.
Future
Political
Situation
−1.01, p < .90
−1.22, p < .90
−1.60, p < .90
−1.97, p < .90
−1.72, p < .90
−2.39, p < .90
−2.76, p < .10
−2.56, p < .90
−1.62, p < .90
−2.58, p < .90
.03, p < 1
2.24, p = .01
0.884 (.17)
.00, p < 1
2.22, p = .01
.909 (.17)
0.344 (.17)
.02, p < 1
2.19, p = .01
0.977 (.17)
0.509 (.17)
.01, p < 1
2.04, p = .02
0.505 (.17)
0.629 (.17)
.00, p < 1
2.23, p = .01
0.553 (.17)
1.001 (.17)
B. Under Putin (January 2000–November 2007 or March 2008, Bimonthly Data)
Putin
Approval
ADF test of I(1)
Phillips-Perron test
of I(1)
KPSS test of I(0)
HML test of I(0)
Estimate of d
Estimate of d for
residuals of
regression of Putin
approval on this
variable
Current
Economy
Family
Finances
Russia’s Ec.
Future
Military Op.
in Chechnya
−1.94, p < .90 −.24, p < .95 −1.09, p < .90 −2.41, p < .90 −2.73, p < .10
−2.78, p < .10 −.68, p < .90 −1.80, p < .90 −3.10, p < .05 −2.81, p < .10
.00, p < 1
2.49, p = .01
.646 (.17)
.14, p < 1
2.35, p = .01
0.725 (.17)
.703 (.17)
.07, p < 1
2.42, p = .01
0.470 (.17)
.754 (.17)
.15, p < 1
−.17, p = .57
0.635 (.17)
.492 (.17)
.17, p < 1
2.28, p = .01
0.604 (.17)
.552 (.17)
Political
Situation
−.44, p < .90
−.95, p < .90
.34, p < 1
2.08, p = .02
0.634 (.17)
.592 (.17)
Note: Calculated using James Davidson’s Time Series Modeling software, v. 4.31; p: probability of the test statistic exceeding the computed
value under H(0); estimates of d calculated with Robinson’s Local Whittle Gaussian ML semi-parametric method; I present averages of the
estimates for bandwidths of 15, 10, and 5 (Yeltsin series have N of 34; those for Putin have N of 50), standard errors in parentheses (averaged
across the three bandwidths). Yeltsin: 10-point rating; Putin: percent approving of his performance. Standard errors in parentheses.
or all series are not stationary and the uncertainty about
whether they are exactly I(1), it makes sense to see whether
they are fractionally integrated.25 I therefore estimated
the order of fractional integration, d, for each series in
each period, using Robinson’s Local Whittle Gaussian
ML semi-parametric method (Robinson 1995; estimates
in Table 1).26 I fractionally differenced each series by its
and possibly also Putin’s approval (Phillips-Perron test marginally
significant).
25
Studies of approval ratings in other countries have also found
them to be fractionally integrated (e.g., Byers, Davidson, and Peel
(2000), which gives theoretical reasons to expect such data to be
fractionally integrated).
26
I used James Davidson’s Time Series Modeling (TSM) software,
v. 4.31. To calculate d, it was necessary to choose a bandwidth
d, estimated for the appropriate period, before including
it in regressions.27
parameter. Unfortunately, as one text puts it: “In the case of the
Gaussian semiparametric estimator. . . [t]here are as yet no satisfactory methods for choosing the bandwidth parameter” (Doukhan,
Oppenheim, and Taqqu 2003, 282). The only recommendation I
could find was that of Haldrup and Nielsen (2007), who suggest
choosing relatively low bandwidths, which tend to “bias estimators
less when noise is not too persistent.” I therefore aimed low, calculating d for bandwidths of 5, 10, and 15, and averaging the results.
N was 34 for the Yeltsin series and 50 for the Putin series.
27
Again, I used Davidson’s TSM software. The fractional differencing formula produces extreme values for the first numbers in
the series (in fact, the first value is the level, not a difference). To
prevent such outliers from distorting the results, I dropped the first
case from the fractionally differenced series for Putin’s approval.
597
PRESIDENTIAL POPULARITY IN RUSSIA
I then examined whether the presidents’ ratings were
cointegrated with any of the economic series. If two variables are cointegrated, a long-run equilibrium relationship exists between them (Box-Steffensmeier and Tomlinson 2000, 70). The usual test for cointegration of two I(1)
variables is to regress one on the other and test whether
the residuals are I(0), in which case the variables are cointegrated. To test whether two fractionally integrated series are fractionally cointegrated, one regresses one on
the other (both in levels) and estimates d for the residuals. If the residuals’ d is less than those for the “parent” series, the variables are fractionally cointegrated.28
The estimates in Table 1 suggest that Yeltsin’s rating was
fractionally cointegrated with perceptions of the current
economy and/or family finances (the two are highly correlated). Putin’s approval may be fractionally cointegrated
with economic expectations and support for continuing the military operation in Chechnya; it may also be
cointegrated with the political mood. However, this was
highly correlated with the economic variables and preferred Chechen policy, and tests in the web appendix
suggest it may have been Granger caused by economic
expectations.
One revealing way to analyze nonstationary time series is with an error correction model, which simultaneously estimates the long-run relationship and the shortrun dynamics of adjustment. Error correction models
have been used with fractionally integrated series to study
various problems (Baum and Barkoulas 2006; Clarke and
Lebo 2003). I use the three-step fractional error correction model of, for instance, Clarke and Lebo (2003) and
Lebo and Cassino (2007). That is, I run regressions of the
form:
d Rati ng t = ␤0 + ␥ j d X j,t + ␦k Wk,t
j
+ ␤1 d E C Mt−1 + εt
k
(1)
where d indicates that a variable has been fractionally
differenced by its estimated value of d; Rating is the average rating on the 10-point scale (for Yeltsin) or the
percentage approving of the president’s performance (for
Putin); the X’s, indexed by j, are fractionally integrated explanatory variables; the W ’s, indexed by k, are stationary
explanatory variables; d E C Mt−1 is the fractional error
correction mechanism (FECM); and ε is a normally dis-
In the Yeltsin period, I fractionally differenced the series including
one interpolated observation from before the regular series started.
I then dropped the corresponding fractionally differenced term.
28
Box-Steffensmeier and Tomlinson (2000) derive this method
from Cheung and Lai (1993) and Dueker and Startz (1998).
tributed stochastic error.29 Where necessary to reduce
autocorrelation, I included one lag of the fractionally
differenced dependent variable on the right-hand side
(Table 3, columns 1, 2, and 4). To obtain the FECM, I
regressed the president’s rating on a right-hand variable
or variables with which it was thought to be cointegrated
(both in levels), estimated d for the residuals from this
regression, fractionally differenced the residuals by this d,
and then lagged the series by one two-month period (see
Clarke and Lebo 2003).
The different economic perceptions measures are
highly correlated (r = .89, for the current economy and
family finances under Putin), which is not surprising since
the same factors are bound to affect all three. Lacking good
instruments, it is easier to assess the aggregate impact of
economic perceptions than to be sure which type matters most. Because of the high correlations, I first present
models with each economic perceptions measure separately and then report a preferred model with more than
one, dropping those event dummies that were not significant at p = .30 (Table 2, column 5; Table 3, column
7). Under Putin, the choice of which economic variables
to put in the final model was somewhat arbitrary given
the high correlations. Including family finances together
with perceptions of the national economy, the former
had an odd negative coefficient, which I judged to be
a spurious result of the correlation between these two
variables rather than evidence that worsening finances
boosted Putin’s popularity. In my view, the data are just
not fine-grained enough to adjudicate between the three
types of perceptions. I also include the political mood
variable separately as it is so highly correlated with the
economic variables and likely Granger caused by one of
them (see the web appendix). I show separate models
with the measures of Putin’s policy performance (on international affairs, order) since these require shortening
the data series. Finally, in Table 2, column 6, and Table 3,
column 8, I show the preferred models with the FECM
dropped; these are needed for subsequent simulations.
Choosing how to model the influence of discrete
events poses a dilemma when the duration of such effects is unclear. One can arbitrarily assume a path of
decay. Or, lacking theoretically informed priors, one can
try several specifications. The second course reduces the
danger of missing the true effect because one has misspecified the duration, but increases the risk of false positives.
In my experience, readers are more uncomfortable with
the latter than with the former. In general, I model each
event with a dummy valued 1 in the month of the earliest
29
The three-step method also avoids including nonstationary variables on the right-hand side.
598
DANIEL TREISMAN
TABLE 2 Explaining Yeltsin’s Popularity, Russia 1994–99
Dependent Variable Is Fractionally Differenced Yeltsin Rating on 10-Point Scale
(d = .884)
(1)
d Current economy
(2)
(3)
.031∗
(.005)
.010
(.012)
d Family finances
First Chechen war start
First Chechen war end
Budyonnovsk crisis
Start of second Chechen war
Drunk in Berlin
Yeltsin hospitalized
Kosovo bombing
1998 financial crisis
Constant
R2
LM autocorrelation test, ␹ 2
KPSS test of I(0)
Durbin Watson statistic
N
(6)
.033∗
(.004)
.005
(.004)
.035∗
(.004)
.007
(.003)
−.005∗
(.002)
.004
(.010)
d Political mood
Months in office
(5)
.011∗
(.003)
d Russia’s economic future
d ECM(t−1)
(4)
−.31∗
(.14)
−.006∗
(.002)
−.17
(.13)
−.004
(.003)
−.004
(.003)
−.004
(.004)
−.23∗
(.09)
−.006∗
(.002)
−.31∗
(.08)
.35∗
(.14)
−.30∗
(.12)
−.55∗
(.09)
.45∗
(.11)
−.22
(.12)
−.41∗
(.09)
.26∗
(.10)
−.14
(.11)
−.51∗
(.18)
.35∗
(.17)
−.16
(.11)
−.28∗
(.06)
.31∗
(.12)
−.25∗
(.06)
−.28∗
(.06)
.25
(.14)
−.16∗
(.05)
−.01
(.07)
−.19∗
(.09)
.03
(.08)
−.05
(.10)
−.19
(.10)
−.08
(.10)
.05
(.10)
−.17
(.09)
−.07
(.09)
.03
(.14)
−.17
(.11)
−.11
(.10)
−.23∗
(.07)
−.27∗
(.07)
.06
(.07)
−.01
(.14)
.16∗
(.07)
.19∗
(.09)
−.14
(.11)
.42∗
(.17)
.8180
.43, p = .51
.05, p < 1
2.00
33
−.53∗
(.19)
.27
(.22)
.6171
.47, p = .49
.08, p < 1
1.73
33
−.52∗
(.09)
.36
(.19)
.6714
.00, p = .96
.13, p < 1
1.91
33
−.59∗
(.16)
.31
(.31)
.5788
.23, p = .63
.13, p < 1
1.73
33
.43∗
(.13)
.8250
.00, p = .97
.06, p < 1
1.83
33
.40∗
(.12)
.7909
.03, p = .85
.07, p < 1
1.97
33
Note:∗ p < .05. OLS with robust standard errors in parentheses. Data are bimonthly, starting in May 1994, when continuous bimonthly
economic series begin, ending in December 1999. d series fractionally differenced using d estimated in Table 1. d E C M(t − 1) in
columns 1 and 5 is the fractionally differenced residuals from a regression of Yeltsin’s rating on perceptions of the current economy (both
in levels), lagged one two-month period; in column 2, the residuals are from a regression of Yeltsin’s rating on perceived family finances.
Column 5 shows the preferred model (in bold), from which event variables not significant at p = .30 have been dropped. Column 6 shows
a regression identical to that in column 5 except without the ECM.
599
PRESIDENTIAL POPULARITY IN RUSSIA
TABLE 3 Explaining Putin’s Popularity, Russia 2000–2007
Dependent Variable Is Fractionally Differenced Percent Approving of Putin’s
Performance (d = .646)
(1)
Current economy
d
(2)
(3)
(4)
(5)
.38
(.11)
d Family finances
−.01
(.23)
d Russia’s ec. future
.18∗
(.07)
(8)
∗
.33
(.16)
.35∗
(.16)
.07
(.08)
.04
(.10)
53∗
(.07)
−.32∗
(.10)
−.04∗
(.02)
.56∗
(.08)
−.05∗
(.02)
−6.73∗
(1.17)
−6.38∗
(1.21)
8.78∗
(.84)
−6.12∗
(1.01)
−3.59∗
(.66)
−3.95∗
(.94)
9.18∗
(1.20)
−5.89∗
(1.18)
−3.59∗
(.75)
−3.12∗
(.84)
6.90∗
(1.08)
8.67∗
(.89)
.19
(.14)
d Successful: international affairs
d Successful: order
.41∗
(.16)
d
−.41∗
ECM(t−1)
(.10)
Months in office
−.03
(.02)
Nordost theater siege
1.12
(2.97)
Beslan terrorist attack
−5.80∗
(1.97)
Sinking of Kursk
−1.05
(1.07)
Soviet anthem restored
8.66∗
(1.16)
9/11
−4.79∗
(1.47)
Iraq war
−3.64∗
(.76)
Monetization of benefits, Jan. 2005 −4.76∗
(1.00)
Takeover of NTV
.72
(1.12)
Arrest of Khodorkovsky
6.50∗
(1.12)
.10
L1 d rating
(.06)
Constant
5.61∗
(1.33)
.7707
R2
2
.18,
LM autocorrelation test, ␹
p = .68
KPSS test of I(0)
.24, p < 1
Durbin Watson statistic
1.86
N
47
(7)
.10
(.07)
d Political mood
d Continue milit. op. in Chechnya
(6)
∗
.31
(.21)
−.41∗
(.13)
−.03
(.02)
2.07
(3.96)
−6.85∗
(2.14)
−2.89
(1.53)
8.64∗
(1.65)
−1.49
(1.91)
−4.94∗
(1.50)
−6.32
(4.56)
.16
(1.46)
4.85∗
(1.40)
.13
(.08)
6.10∗
(1.53)
.6887
.00,
p = .96
.24, p < 1
1.94
47
.45∗
(.12)
−.34∗
(.11)
−.04
(.02)
1.69
(2.45)
−6.79∗
(1.76)
−1.94
(1.37)
7.95∗
(1.17)
−4.24∗
(.69)
−4.44∗
(.80)
−4.30∗
(1.05)
.02
(1.37)
5.45∗
(.99)
.32
(.21)
−.46∗
(.13)
−.03
(.02)
3.77
(4.41)
−2.97
(3.96)
−1.63
(1.63)
9.08∗
(1.45)
−2.18
(1.62)
−4.60∗
(.77)
−5.98∗
(1.02)
.33
(1.36)
4.83∗
(1.22)
.13
(.08)
7.12∗
5.50∗
(1.39)
(1.74)
.7118
.7054
.35,
.06,
p = .55 p = .80
.31, p < 1 .23, p < 1
1.81
1.89
47
47
.41∗
(.17)
−.13
(.13)
−.05
(.03)
−1.93
(3.19)
−6.25∗
(2.00)
−5.05∗
(2.21)
10.34∗
(1.41)
−2.82
(1.98)
−4.31∗
(.82)
−4.26∗
(1.15)
−2.25
(2.85)
5.01∗
(2.14)
.11
(.22)
.40∗
(.19)
−.14
(.14)
−.04
(.03)
.22
(3.43)
−7.13∗
(1.85)
−2.22
(2.71)
10.04∗
(1.85)
−1.85
(1.53)
−5.47∗
(1.93)
−4.71∗
(1.49)
.19
(1.62)
7.10∗
(2.03)
5.82∗
7.11∗
6.74∗
6.97∗
(1.53)
(1.78)
(.95)
(1.25)
.6420
.5984
.7540
.6936
.16,
.15,
.33,
.79,
p = .69 p = .70 p = .57
p = .38
.14, p < 1 .22, p <1 .29, p < 1 .24, p < 1
2.08
1.87
1.81
2.19
40
40
47
47
Note:∗ p < .05. OLS with robust standard errors in parentheses. Data bimonthly, starting in early 2000 and ending in December 2007. d
series fractionally differenced using d estimated in Table 1. d E C M(t − 1) is the fractionally differenced residuals from a regression of
Putin’s approval on expected performance of the national economy (both in levels), lagged one two-month period. Preferred model in
bold.
600
subsequent survey, and 0 at other times. The impact is
thus assumed to decay at the same gradual rate that the
fractional differencing implies.
The one exception is the end of the first Chechen war,
which arrived gradually. In March 1996, Yeltsin decreed a
halt to military operations, but this changed little on the
ground. In May 1996, he and acting Chechen President
Yandarbiyev signed a cease-fire agreement. But only in
late August was the Khasavyurt Accord signed, actually
ending the war. I model this with a dummy valued 1 in
May and July, and 0 otherwise (June and August were
not in the bimonthly data). I include simple dummies for
the months in which the two Chechen wars started (since
the dependent variables are differenced, dummies for the
start and end make more sense than a dummy for the
duration).
As is common in studies of U.S. presidential approval,
I control for the time the president had been in office. Approval of American presidents typically falls, even controlling for other factors. Mueller (1973) argued that citizens
start with unrealistically high expectations and gradually
grow disappointed in their leaders. Under Putin, months
in office and the index of press restrictions turn out to be
highly correlated (r = .96); both rose monotonically over
time. Therefore, I do not include them in the same regression; months in office turned out to be more significant
in each case.
Table 1 suggested that, under both presidents, more
than one variable might be cointegrated with the president’s rating. Since it was not obvious a priori which of
these—individually or in combination—belonged in the
FECM, I tried several formulations. In Table 2, columns 1
and 5 contain an FECM formed using the residuals from
regressing Yeltsin’s rating on perceptions of the current
economy. In column 2, the FECM used residuals from
Yeltsin’s rating regressed on perceived family finances.
The former proved much more significant, and also more
significant than an FECM for which the cointegrating
regression contained both family finances and the current economy. In Table 3, the FECM uses residuals from
regressing Putin’s approval on economic expectations.
This proved more significant than FECMs that incorporated preferences on Chechnya policy or the political
mood.
What do the results show? Consider first the effects of
war and terrorist attacks. Yeltsin’s Chechen misadventure
appears to have cost him support, although perhaps less
than might have been expected. His rating fell at the start
of the first war by more than one quarter of a point on
the 10-point scale (Table 2, column 5). It rose as the
war ended in 1996, by .31 of a point in both May and
July; of course, the estimates should be viewed as only
DANIEL TREISMAN
approximate. Yeltsin lost another quarter point after the
1995 Budyonnovsk terrorist siege. By contrast, the start
of the second war had no clear effect.30
Putin’s approval rose and fell with support for the
use of military force in Chechnya. Since the latter peaked
at 70% in March 2000, falling to just 13% in December
2007, this suggests the Chechen situation helped Putin
early on, but weighed him down in later years. Indeed,
supposing the coefficients on the fractionally differenced
series roughly correspond to those for first differences,
Putin lost about 20–30 points of approval over the course
of his presidencies as faith in his military approach dwindled. Terrorist attacks temporarily revived backing for
military force. After the Nordost siege in 2002, support
for the military option leapt from 34% in September to
48% in November, before falling back to 31% in January
2003. This led to a rally behind Putin, whose approval
temporarily jumped by six points. (Nordost had no effect beyond its militarization of the public mood; the
dummy is insignificant in Table 3, where regressions control for support for the military operation.) Beslan led to
a smaller jump in support for military force (from 24 to
31%); but Putin’s approval actually fell that month. The
Beslan dummy is significantly negative, controlling for
preferences on Chechnya policy; in other words, support
for Putin rose by less than one would have expected given
the militarization of the public mood. One might speculate that in this case, Russians blamed both the terrorists
and the authorities for the disastrous outcome.
I lacked good data to assess the influence of presidential style conclusively. But it probably made some
difference. After Yeltsin conducted the band in Berlin, his
rating fell by about one quarter point. I found no evidence,
however, that Yeltsin’s hospitalizations hurt his popularity. Perhaps his ill health was already too well known for
this to matter. Putin’s approval may have fallen after the
Kursk sank, but the estimates were statistically insignificant. Restoring the Soviet music to the national anthem
produced a boost of eight or nine points. It could be that
this was the moment Putin “closed the deal” with some
former Communist supporters. Perhaps for the same reason, the arrest of the oligarch Khodorkovsky led to a
bounce of about seven points.
Neither respondents’ assessments of the political
situation nor their evaluations of Putin’s success in
30
That the dependent variable is fractionally differenced makes it
harder to interpret the size of the effects. To get a sense of what
difference this makes, I ran versions of the preferred models firstdifferencing instead of fractionally differencing the dependent variables. The coefficients were mostly quite similar, at least for the
significant variables, so it is probably safe to assume the effects are
roughly as large as those in Tables 2 and 3.
601
PRESIDENTIAL POPULARITY IN RUSSIA
international affairs or at establishing “order” significantly affected presidential ratings, although in the latter cases the limited data might be to blame. After the
United States invaded Iraq, Putin’s approval fell several
points: apparently the exercise of American power cast
his previous solicitousness of the United States after 9/11
in a bad light. Putin’s rating also fell after the 9/11 attack. NATO’s bombing of Kosovo did not have a clear
impact on Yeltsin’s rating (it was not significant in the
final model).
As with U.S. presidents, a time trend remains in
Yeltsin’s and Putin’s ratings controlling for other variables. One should interpret the months’ term along with
the intercept; the first was negative for both presidents,
the second positive. This implies that, other things equal,
approval tended to rise at first, but then fall. Such a pattern makes intuitive sense: at first citizens rallied behind
a new leader, but over time more and more grew disillusioned for a variety of reasons not captured by the
model. For Yeltsin, the effect turned negative after about
70 months (mid-1997); for Putin, it remained positive
through the end of his second term.31 The admittedly
rough press freedom data do not suggest greater Kremlin control boosted Putin’s rating. The NTV takeover
had no clear effect. Nor did Freedom House’s media
index; always less significant than time in office, it was
dropped from the regressions. Months in office might,
under Putin, be picking up the trend toward less press
freedom. But if so this would suggest that unfree media
lowered Putin’s approval: the coefficient on months was
negative.
These results include a mixture of surprises and confirmations of the conventional wisdom. Some of the hypothesized factors help account for the spikes in the ratings at various points. But most effects are small. By contrast, economic perceptions have considerable explanatory power.32
Under Yeltsin, perceptions of the national economy had the strongest impact. The significant FECM in
columns 1 and 5 suggests a long-run relationship between rosy views of current conditions and support for
the president. The coefficient’s value, –.23 in column 5,
implies that when a shock knocks the two variables out
of equilibrium, about one quarter of the gap is closed
each period. The significant coefficient on fractionally
differenced current economic perceptions suggests these
also had a short-term effect. As expected, the 1998 financial crisis depressed Yeltsin’s popularity. This is not
significant if one controls for perceptions of the current economy, suggesting the crisis affected Yeltsin’s rating via this pathway. Under Putin, the evidence suggests
a long-run relationship between economic expectations
and presidential approval, with about one-third of any
divergence eliminated each period (coefficient of –.32 in
column 7). I also found evidence of short-run effects, although the correlations among variables make it hard to
say which matters more. Current economic assessments
and expectations are both significant if entered alone
(columns 1 and 3), but statistical significance falls if more
than one is included at one time. Again, the strongest
short-run effects are of perceptions of the current economy. As expected, Putin’s monetization of benefits cost
him several percentage points even after taking into account the impact of the reform on perceptions of family
finances.
How much difference did the economy make? One
can explore this by simulating the final model estimated for one president, substituting economic perceptions from the corresponding month in the other’s term,
while leaving all other variables at their actual levels.
The two presidents’ simulated ratings supposing each had
presided over the other’s perceived economic conditions
are shown in Figure 2.33 This exercise should be taken
with a grain of salt. The simulations change somewhat
depending on the specification, and it was not possible to include an FECM.34 Nevertheless, the results are
suggestive. It seems Russians’ radically different evaluations of their first two presidents were strongly influenced
by the very different economic conditions under which
each served. Had Yeltsin presided over Putin’s economy,
he would apparently have left office extremely popular.
Had Putin presided over Yeltsin’s, his rating would have
plunged early on, recovered a bit, but then sunk again.
In January 2008, fewer than 50% of Russians would have
approved of his performance.35
33
31
To locate the peak, I divide the constant by the coefficient on
months in the preferred models.
32
For instance, a model including all the variables in Table 2 except
the economic ones has an adjusted R2 of .1270. Adding fractionally
differenced current economy and family finances along with the
FECM raises the adjusted R2 to .7036. The difference is less dramatic
but still notable in the Putin period—adding economic variables
to a model with all noneconomic variables raises the adjusted R2
from .4907 to .6630.
To simulate Putin with Yeltsin’s economy, it was necessary to
impute values for economic perceptions under Yeltsin in 1992. To
do this, I regressed the economic perceptions variables on inflation
and real wages during months when both were available and used
the resulting models to impute backwards.
34
Since one is simulating the dependent variable, one cannot first
regress it on economic perceptions to form the FECM. Instead, I
use the models in Table 2, column 6, and Table 3, column 8.
35
In fact, this probably underestimates Putin’s decline. Not including the FECM, the simulations are unable to incorporate the
602
DANIEL TREISMAN
FIGURE 2 Simulating Yeltsin’s Rating with Putin’s Economy (top);
Simulating Putin’s Approval with Yeltsin’s Economy
(bottom)
The regressions also enable us to explore an intriguing counterfactual about Putin’s rise, which is usually seen
as inexorably tied to the violent events of late 1999. Suppose there had been no apartment bombings, no second
Chechen war. Would Putin’s rating have stayed around
its starting point of 31%? Of course, we cannot know
for sure, but the statistics offer some clues. To see what
economic perceptions alone would predict, I used the estimated effects of just economic perceptions and months
in office from Table 1, Model 6, and simply extrapolated
forward using the actual economic perceptions data under Putin and restarting months in office. In Figure 3, the
long-run effect of the fall in economic perceptions under Yeltsin’s
economy.
dotted line shows the simulated rating of a generic new
president based on just the dramatic economic recovery,
supposing Russians started evaluating the new leader in
September 1999, right after Putin became prime minister. The dashed line supposes Russians started evaluating
the new incumbent in January 2000, when Putin became
acting president.36
36
To be clear, no actual data on Putin’s approval were used to calculate this. I simply used actual economic perceptions to simulate
the incumbent’s rating, restarting months in office in the relevant
month. I therefore assume nothing about Yeltsin’s successor except
that he took office in September 1999 (dotted line) or January 2000
(dashed line) and that economic perceptions influenced his rating
in the same way as under Yeltsin.
603
PRESIDENTIAL POPULARITY IN RUSSIA
FIGURE 3 Predicting Putin’s Approval Rating with Economic
Perceptions
Sources: VCIOM/Levada Center and author’s calculations.
The striking implication is that even without a
Chechen war or terrorist attacks, economic factors alone
would have prompted a leap in the new president’s popularity quite similar to that which occurred. In late 1999,
the economy began to revive. Between August 1999 and
June 2000, real wages rose by about 20%, while wage arrears and unemployment fell. This fueled an unexpected
rebirth of economic optimism that would have boosted
support for any new Kremlin incumbent.37
The main difference is that the surge in approval predicted by economic factors comes a little later than the
actual surge. In those months, time was of the essence.
Putin would have to run for election in June 2000, or,
after Yeltsin resigned early, in March. Were it not for the
Chechen war, the Kremlin would have been best served
by the original election schedule. If the public began evaluating Putin in September 1999, by July 2000 the simulation suggests his rating would have reached 4.5 on the
10-point scale. Supposing instead that Putin remained
tethered to Yeltsin’s rating until January 2000, by July his
would have been 3.7. In 1996, when Yeltsin won reelection
with 54% of the valid vote, his rating had been 3.9. Thus,
it is very possible that Putin—or some other new Kremlin
candidate—would have won the presidency even without
the Chechen conflict and terrorist bombings.
37
Had Yeltsin been eligible to run for a third term and continued in
office, simulations suggest the economic recovery would also have
revived his popularity—but more slowly and less dramatically. As
noted, the accumulated effect of time in office was by that point
weighing quite heavily on his rating.
However, economic factors do not explain why
Putin’s rating actually surged when it did. This might,
indeed, be associated with the Chechen events. Although
economic factors would have achieved the same result
a few months later, the traumas of late 1999 apparently
sped things up.38
The Determinants of Economic
Perceptions
If economic perceptions were important, what caused
these perceptions? Did they reflect actual economic conditions, were they idiosyncratic, or were they manipulated by government propaganda? Russians’ rosy view of
the economy under Putin might itself result from the
Kremlin’s media control. Or positive economic perceptions might reflect a general confidence in Putin’s stewardship.
To explore this, I ran fractional error correction models with the fractionally differenced economic perceptions
measures as dependent variables (Table 4). Among explanatory factors, I included six objective measures of
economic conditions—the average real wage, real wage
arrears, the average real pension, logged inflation, unemployment, and the demand for workers (job openings
38
As noted, support for military operations fell drastically in subsequent years, which must have pushed down Putin’s rating; that it
remained high reflects the positive effect of economic stabilization.
604
DANIEL TREISMAN
TABLE 4 Determinants of Economic Perceptions
d Current Economy
(d = .947)
Dependent Variable
Economic data
d Real wage
d Real wage arrears
d Average real pension
d Log inflation
d Unemployment
d Demand for workers
d ECM(t−1)
Recovery phase
Economic events
1998 financial crisis
Oct. 1994 financial crisis
Monetization of benefits, Jan. 2005
Politics
Putin presidency dummy
President’s rating (10-point scale)
Khodorkovsky arrested
Press freedom (high = less free)
1996 campaign
2000 campaign
2004 campaign
2008 campaign
(1)
.007
(.011)
−.041
(.035)
.023
(.023)
4.24
(3.73)
−.21
(.98)
.007
(.006)
−.50∗
(.08)
.97
(1.73)
−20.16∗
(4.91)
−6.43∗
(2.48)
−.48
(1.33)
.41
(1.71)
3.08∗
(1.08)
−6.68∗
(1.45)
.00
(.09)
3.95∗
(1.38)
−1.30
(2.37)
5.37∗
(1.08)
−1.75
(1.58)
(2)
−.040
(.024)
.008
(.004)
−.51∗
(.08)
1.75∗
(.82)
−18.84∗
(2.39)
−5.76∗
(.99)
2.78∗
(.83)
−6.14∗
(1.17)
4.13∗
(1.03)
5.41∗
(1.02)
d Family Finances
(d = .832)
(3)
.004
(.011)
.073
(.057)
.051∗
(.022)
2.48
(4.23)
−2.49
(1.34)
−.001
(.005)
−.48∗
(.09)
8.66∗
(2.27)
−.61
(6.97)
−1.65
(2.58)
−15.27∗
(2.01)
−4.99
(2.50)
2.11∗
(.72)
.45
(1.98)
−.06
(.08)
−1.19
(1.38)
−1.10
(1.99)
−.17
(1.10)
−3.66∗
(1.42)
(4)
.012
(.007)
.051∗
(.019)
−2.82∗
(1.33)
−.45∗
(.11)
3.13∗
(.89)
−15.05∗
(1.64)
1.44∗
(.45)
−1.70
(.92)
−4.46∗
(.65)
d Russia’s Ec. future
(d = .884)
(5)
.007
(.020)
−.12∗
(.05)
.10
(.05)
3.04
(7.40)
−2.91
(2.40)
−.00
(.01)
−.27
(.15)
1.21
(4.54)
−2.99
(7.60)
−11.73∗
(5.39)
−5.89
(3.06)
−3.62
(4.69)
7.23∗
(1.71)
−6.24∗
(2.80)
.10
(.17)
3.57
(2.15)
6.16
(7.19)
8.53∗
(1.70)
−5.89∗
(2.33)
(6)
−.11∗
(.05)
.094∗
(.035)
−2.77
(1.83)
−.27∗
(.12)
−12.48∗
(1.28)
−5.41∗
(2.05)
6.91∗
(1.69)
−6.10∗
(2.15)
3.86∗
(1.93)
6.94
(7.15)
8.79∗
(1.54)
−4.29∗
(.94)
(Continued)
605
PRESIDENTIAL POPULARITY IN RUSSIA
TABLE 4 Continued
d Current Economy
(d = .947)
Dependent Variable
Constant
R2
LM autocorrelation test, ␹ 2 (p)
KPSS test of I(0)
Durbin Watson statistic
N
d Family Finances
(d = .832)
d Russia’s Ec. Future
(d = .884)
(1)
(2)
(3)
(4)
(5)
(6)
−.53
(4.84)
.6537
.02 (.90)
.04, p < 1
2.00
80
−.63
(.67)
.6255
.19 (.67)
.04, p < 1
2.05
80
.82
(4.50)
.6782
.03 (.86)
.04, p < 1
1.95
80
−2.31∗
(.75)
.6030
.42 (.52)
.19, p < 1
2.06
80
−4.76
(8.75)
.5129
.19 (.66)
.05, p < 1
2.08
80
.31
(.73)
.5039
.20 (.66)
.10, p < 1
2.09
80
Note: ∗ p < .05. OLS with robust standard errors in parentheses. Data bimonthly, starting in early 1994, ending in December 2007. d :
series fractionally differenced using estimated d (average for bandwidths of 30, 20, and 10). d E C M(t − 1): fractionally differenced one
period lag of residuals of regression of dependent variable on (a) real wage and demand for workers (columns 1–2); (b) real wage and
unemployment (columns 3–4); (c) real wage (columns 5–6). Even columns show models from which the most insignificant variables have
been dropped.
reported to the state employment service). As might be
expected, these were correlated (e.g., r = .87 for the real
wage and pension), so besides showing models that include all six (odd-numbered columns), I present simpler models from which the most insignificant variables
have been dropped (even columns). Given the high correlations, conclusions about which economic variables
were most important have to be tentative. All the economic indicators appeared to be fractionally integrated,
so I fractionally differenced each by the appropriate d
(average estimate for bandwidths 10, 20, and 30; N =
82–4) and included a fractional error correction mechanism. Based on exploratory analysis, the most appropriate
FECM was based on cointegrating regressions, including
the real wage (columns 5–6), the real wage and demand
for workers (columns 1–2), and the real wage and unemployment (columns 3–4).
Besides the official statistics, certain events informed
the public about economic conditions. One might expect gloomier views after the August 1998 financial crisis.
Earlier, Russians had been shocked when on “Black Tuesday” in October 1994 the ruble plunged almost 30%,
prompting Yeltsin to fire his main economic ministers.
The protests that followed the January 2005 reform of
social benefits suggest many Russians thought this had
worsened their financial situation. I also include a dummy
for whether the economy was in its decline or recovery
phase, on the theory that expectations tend to overshoot,
being too pessimistic in downturns and too optimistic
in recoveries. It is almost a cliché that in business cycles
consumers swing between “irrational exuberance” in the
boom and excessive pessimism in the recession; there is
some evidence for this from the United States.39
What about political influences? First, causation
might run in reverse from presidential approval to perceptions of economic performance. In the United States,
overall presidential approval does not appear to affect
consumer sentiment (MacKuen, Erikson, and Stimson
1992), but approval of the president’s economic management does (De Boef and Kellstedt 2004). So in estimating
the impact of economic indicators on perceptions, one
should at least control for presidential ratings; I include
the incumbent’s differenced rating on the 10-point scale.
(My estimate of d was almost exactly 1, so first differencing was appropriate.) Certain government actions are
likely to affect economic perceptions. I included a dummy
to see whether Russians viewed Khodorkovsky’s arrest as
good or bad for the economy (recall that Putin’s approval
jumped). Finally, I looked for effects of media manipulation. I included a dummy for Putin’s presidency given
the view that coverage became more favorable under him,
and also included the Freedom House index of press freedom. Media effects are expected to be particularly strong
during election campaigns. I included variables valued 1
for the six months before a presidential election and –1for
the six months after it.40
39
Tortorice (2009), for instance, finds that American poll respondents are irrationally pessimistic about future unemployment at
the end of a recession and excessively optimistic at the beginning of
the downturn. I date the bottom in Russia by the month in which
real wages were lowest, February 1999.
40
For a similar approach, see De Boef and Kellstedt (2004).
606
Table 4 suggests that Russians’ views of economic
conditions were not just idiosyncratic. Each perceptions
variable is systematically related to objective measures of
economic performance. All three appear to be in long-run
equilibrium with the average real wage; perceptions of the
current economy and family finances were apparently also
in long-run equilibrium with, respectively, demand for
workers and unemployment. When real wages rise, economic expectations adjust upward to a new equilibrium,
closing about a quarter of the gap each period (column
6). In the short run, more job vacancies (significant at p
= .06) and perhaps lower wage arrears (p = .09) correlate with a rosier view of the current economy. Short-run
increases in pensions (p = .01) and drops in unemployment (p = .04) were associated with cheerier views of
family finances; higher wages may also have helped (p
= .08). Rising pensions (p = .01), falling wage arrears
(p = .04), and possibly falling unemployment (p = .13)
led to more positive expectations. Inflation was never significant; its effects might be captured here by real wages
and pensions. Attitudes did also overshoot, accentuating
the current trend; even controlling for political factors,
views of the current economy and family finances were
more negative in the downslide and positive during the
recovery than would be predicted from just the objective
indicators.
As expected, major economic events also triggered
reevaluations. The 1998 financial crisis fostered gloom
about the current economy, although oddly it did not influence expectations. Both views of the current economy
and economic expectations worsened after the October
1994 currency crisis. However, these events did not affect
Russians’ assessments of their own finances, unless such
effects were captured by the economic indicators. By contrast, the 2005 monetization of benefits prompted a sharp
fall in how well-off many Russians felt, as well as more
pessimistic expectations.
The extent to which objective economic indicators
and events can, by themselves, account for the pattern of
economic perceptions can be seen in Figures A2.A and
A2.B in the web appendix. In these graphs, I plot both
actual economic perceptions and those predicted using
just the economic variables in the regressions in columns
2 and 4, leaving out political effects. The predictions do
not catch all the spikes and dives in perceptions. But they
do an excellent job of capturing the trends.
Politics helps explain the remaining variation. As expected, presidential approval correlated with all three
economic perceptions measures; controlling for it allows greater confidence in the estimated impact of
the economic variables. The coefficients on presidential
approval—2.78, 1.44, and 6.91, in columns 2, 4, and 6—
DANIEL TREISMAN
suggest the extent of reverse causation. For the current
economy and family finances, such effects seem minor.
Of the 84-point range of the current economy variable,
presidential popularity could predict about 16 points, and
it could predict about eight points of the 77-point range
in family finances.41 The possible effect of presidential
approval on economic expectations is somewhat larger—
the former could predict up to 39 points in the 81-point
range of the latter.
The dummy for Putin’s presidency was not significant in columns 1 or 5. It was significant at p = .051 but
negative in Model 3 for family finances. Of course, it was
highly correlated with the economic recovery variable,
which distinguished months after February 1999. Interpreting the two together suggests that perceived family
finances improved a lot in the early recovery months of
1999, but that the effect diminished after January 2000.
Including the economic recovery and Putin dummies
separately in the preferred family finances model (column 4), economic recovery is highly significant but the
Putin dummy is not. In short, the evidence suggests that
perceptions were better during the recovery phase, which
largely corresponded to Putin’s presidencies. But it is not
clear that Putin’s leadership (and media management)
had an independent effect beyond that already captured
by the presidential rating. Nor was there evidence that
economic perceptions improved as press freedom decreased, at least as measured by the Freedom House index.
Khodorkovsky’s arrest led to a boost of some six or
seven points in Putin’s approval (Table 3). Apparently this
was in spite of—rather than because of—the economic
effect Russians anticipated: after the arrest, assessments
of and expectations for the economy worsened. Russians
may have favored the arrest on grounds of social justice
despite fearing adverse economic effects; or perhaps it was
different Russians who applauded Putin’s action and who
anticipated economic reverberations.
Finally, presidential campaigns do seem at times to
have influenced perceptions. In 1996 and 2004, Russians’
views of the economy and economic expectations improved during the six preelection months and fell during the six postelection months by more than expected
based on the economic data. Interestingly, the campaigns
brought no improvement in Russians’ perceptions of their
own material situations—these may even have worsened
in 1996—so this might indeed reflect the media’s influence. There were no significant campaign effects in 2000,
and November 2007 (the only month in the dataset from
41
The range of the presidential rating was 5.6 points. Multiplying
this by the coefficients on current economy and family finances
yields 15.6 points for current economy and 8.1 points for family
finances.
PRESIDENTIAL POPULARITY IN RUSSIA
the 2008 campaign) actually saw a decline in economic
expectations and perceived family finances. Whether the
Medvedev campaign later kicked into gear must await
future research.42
Thus, politics does help explain some of the peaks
and dips in economic perceptions. Russians’ greater approval of Putin than of Yeltsin may have fed back, exacerbating economic pessimism in the late Yeltsin period
and fueling optimism in the early Putin years. Biased media might help explain the surges and falls of economic
optimism around the presidential campaigns of 1996 and
2004. On the other hand, I found no evidence of systematically greater distortions under Putin than under Yeltsin.
Overall, the trends in economic perceptions were well
predicted by just the economic indicators. Perceptions of
the economy plummeted under Yeltsin and soared under
Putin not because of effective propaganda but because the
underlying economic variables were doing the same.
Conclusion
In illiberal democracies and semi-authoritarian states,
politicians’ ratings wax and wane, sometimes very rapidly.
Electorates swing between candidates and parties, at times
embracing complete unknowns. It is easy to attribute such
fluctuations to an unhealthy brew of rootless and cynical
voters, demagogic politicians, and media manipulation.
But in some cases the cause may be something else: electoral volatility may mirror economic volatility—and the
efforts of rational voters, against the odds, to hold their
leaders to account.
In Russia, Yeltsin’s plunging popularity and Putin’s
soaring ratings are often linked to their personalities and
public images.43 Some see in Putin’s ratings an endorsement of his Chechen military campaign and efforts to
rebuild the Russian state (Sil and Chen 2004). Others suppose Russians backed Putin because they had been brainwashed by a state-controlled media from which criticism
of the Kremlin had been eliminated (Ryzhkov 2004).
Wartime rallies, media effects, and image did play a
part at times. Perhaps better data would show their role to
42
If one assumes each campaign had the same effect, the variable is
significant at p = .01 for the current economy, p = .03 for economic
expectations, and not at all for family finances.
43
Anderson (2007) attributes Putin’s popularity largely to his “image of firm, where necessary ruthless authority,” and claims to see a
link to Russian culture: “Historically, the brutal imposition of order
has been more often admired than feared in Russia. . . . In what remains in many ways a macho society, toughness—prowess in judo
and drops into criminal slang are part of Putin’s kit—continues to
be valued.”
607
be somewhat greater. However, Russians’ perceptions of
economic conditions have been more consistently important. Much like citizens in Western democracies, Russians
approved of their president when they saw the economy
improving and felt hopeful about its future. Although
Kremlin public relations influenced perceptions, especially in certain electoral campaigns, Russians’ views of
the economy were closely related to objective economic
indicators such as real wages, pensions, and wage arrears.
From this perspective, the fates of Russia’s first two
presidents appear in a new light. Yeltsin’s ignominious
slide and Putin’s eight years of adulation seem largely
predetermined by the economic conditions each inherited. (Of course, each influenced the economy on his
watch; still, Yeltsin inherited the Soviet system in midcollapse, whereas Putin benefited from a recovery fueled
by soaring oil prices.) Simulations suggest a generic new
president would have become extremely popular—judo
or no judo—as a result of the boom.
Starting in late 2008 (after the first draft of this
article was written), Russia succumbed to the international financial crisis. Yet the ratings of Putin (now prime
minister) and Medvedev, his successor, did not drop
dramatically. Has the relationship between economics
and presidential popularity in Russia changed? In fact,
considered more closely, recent data offer a striking
confirmation of the relationships noted here. Two circumstances are important. First, the Putin government,
concerned about social stability, succeeded in sheltering
the public from the full pain of the crisis. Despite an 8%
drop in GDP per capita in 2009, real wages fell only 2.8%;
real disposable incomes actually rose by 1.9% because of
generous increases in pensions, which rose by 10.7% in
real terms that year. Unemployment only increased from
5.7% in July 2008 to 8.2% in December 2009, less than the
United States’s 9.7%. Although alarming, the economic
downturn was relatively mild in its effects. Second, the
effect of deteriorating economic sentiment was largely
offset by a surge in support for Putin and Medvedev during the war with Georgia in August 2008. Between July
and September 2008, Putin’s approval leapt from 80.4%
to 88.0% and Medvedev’s rose from 69.4% to 83.0%. This
was a classic wartime rally behind the flag.
If one subtracts out the 7.6-point jump in Putin’s
approval in September 2008 attributable to the Georgian
war, his rating follows almost exactly the course predicted
by the model I previously estimated for the Putin presidency (column 8 in Table 3) when one enters the actual
economic perceptions data for the Medvedev presidency
(see Figure A3 in the web appendix). The percentage approving of Putin’s performance falls in the simulation
from 85 in March 2008 to 69 in November 2009; Putin’s
608
DANIEL TREISMAN
actual rating minus the Georgia jump fell from 85% to
71%. Medvedev’s rating correlates with Putin’s at r = .85
in this period; if we subtract out his even larger jump
during the Georgian war, his approval closely parallels
the predicted trajectory.44 Thus, the financial crisis of
2008–2009 does appear to have pulled down the leaders’
popularity, but by a fairly moderate amount since the government took active measures to shelter the population,
and in a way that was largely offset by the wartime rally of
support for the Kremlin over the conflict with Georgia.
The findings of this article fit with a body of recent
work that has been discovering familiar, rational behavior
beneath the irregular surfaces of political life in developing and middle-income countries (Drazen 2008). Outside
the rich democracies, information is usually asymmetric, uncertainty is endemic, and economic conditions often fluctuate wildly. In such environments, quite rational
behavior can look like impulsiveness and manipulation.
While miscalculations and fraud are certainly common in
the hybrid regimes of the developing and postcommunist
worlds, so too, it turns out, is retrospective voting and
economics-based evaluations of incumbents. From Peru
to Zambia, “governments are being held accountable for
bad economic policies, at least to some degree” (LewisBeck and Stegmaier 2008). The irony is that in such countries economic conditions are particularly vulnerable to
global forces, which complicates the task for citizens—
even sophisticated ones—of separating noise from signals
about leaders’ competence.
References
Anderson, Perry. 2007. “Russia’s Managed Democracy.” London
Review of Books, January 25, 2007, 3–12.
Arce, Moisés. 2003. “Political Violence and Presidential Approval in Peru.” Journal of Politics 65 (May): 572–83.
Baum, Christopher, and John Barkoulas. 2006. “Dynamics of
Intra-EMS Interest Rate Linkages.” Journal of Money, Credit,
and Banking 38: 469–82.
Box-Steffensmeier, Janet M., and Renée M. Smith. 1996. “The
Dynamics of Aggregate Partisanship.” American Political Science Review 90: 567–80.
44
Since the Levada Center has not published regular data on economic perceptions for this period, I had to interpolate some observations either linearly or, where possible, regressing the Levada
data on answers to identical questions asked in surveys by VCIOM,
the polling organization Levada used to head. (The results of economic questions on surveys by the two organizations are actually
very similar.) Since the Levada Center stopped asking the question
about strategy in Chechnya in December 2007, I assumed support
for the military option remained at its December 2007 level of 13%.
Box-Steffensmeier, Janet M., and Andrew R. Tomlinson. 2000.
“Fractional Integration Methods in Political Science.” Electoral Studies 19: 63–76.
Byers, David, James Davidson, and David Peel. 2000. “Modeling
Political Popularity: An Analysis of Long-Range Dependence
in Opinion Poll Series.” Journal of the Royal Statistical Society:
A 160: 471–90.
Cheung, Yin-Wong, and Kon S. Lai. 1993. “A Fractional Cointegration Analysis of Purchasing Power Parity.” Journal of
Business and Economic Statistics 11: 103–12.
Clarke, Harold D., and Matthew Lebo. 2003. “Fractional
(Co)integration and Governing Party Support in Britain.”
British Journal of Political Science 33: 283–301.
Clarke, Harold, and Marianne C. Stewart. 1995. “Economic
Evaluations, Prime Ministerial Approval and Governing
Party Support: Rival Models Reconsidered.” British Journal
of Political Science 25: 145–70.
Colton, Timothy J. 2000. Transitional Citizens: Voters and What
Influences Them in the New Russia. Cambridge, MA: Harvard
University Press.
Colton, Timothy J. 2008. Yeltsin: A Life. New York: Basic Books.
Colton, Timothy J., and Henry E. Hale. 2008. “The Putin Vote:
The Demand Side of Hybrid Regime Politics.” George Washington University. Unpublished manuscript.
De Boef, Suzanna, and Paul M. Kellstedt. 2004. “The Political (and Economic) Origins of Consumer Confidence.”
American Journal of Political Science 48: 633–49.
Dominguez, Jorge I., and James A. McCann. 1996. Democratizing Mexico: Public Opinion and Electoral Choices. Baltimore:
Johns Hopkins University Press.
Doukhan, Paul, Georges Oppenheim, and Murad S. Taqqu, eds.
2003. Theory and Applications of Long-Range Dependence:
Theory and Applications. Basel: Birkhäuser.
Drazen, Allan. 2008. “Is There a Different Political Economy for
Developing Countries? Issues, Perspectives, and Methodology.” Journal of African Economies 17: 18–71.
Duch, Raymond M. 1995. “Economic Chaos and the Fragility
of Democratic Transitions in Former Communist Regimes.”
Journal of Politics 37: 121–58.
Dueker, Michael, and Richard Startz. 1998. “MaximumLikelihood Estimation of Fractional Cointegration with an
Application to U.S. and Canadian Bond Rates.” Review of
Economics and Statistics 80 (August): 420–26.
Ekman, Joakim. 2009. “Political Participation and Regime Stability: A Framework for Analyzing Hybrid Regimes.” International Political Science Review 30: 7–31.
Erikson, Robert S., Michael B. MacKuen, and James A. Stimson.
2002. The Macro Polity. New York: Cambridge University
Press.
Haldrup, Niels, and Morten Orregaard Nielsen. 2007. “Estimation of Fractional Integration in the Presence of Data Noise.”
Computational Statistics & Data Analysis 51: 3100–3114.
Hale, Henry. 2009. “The Myth of Mass Authoritarianism in
Russia: Public Opinion Foundations of a Hybrid Regime.”
George Washington University. Unpublished manuscript.
Hesli, Vicki L., and Elena Bashkirova. 2001. “The Impact of
Time and Economic Circumstances on Popular Evaluations
of Russia’s President.” International Political Science Review
22: 379–98.
PRESIDENTIAL POPULARITY IN RUSSIA
Jeffries, Ian. 2002. The New Russia: A Handbook. New York:
Routledge.
Kramer, Gerald. 1983. “The Ecological Fallacy Revisited:
Aggregate- versus Individual-level Findings on Economics
and Elections, and Sociotropic Voting.” American Political
Science Review 77: 92–111.
Lafay, Jean-Dominique. 1991. “Political Dyarchy and Popularity Functions: Lessons from the 1986 French Experience.”
In Economics and Politics: The Calculus of Support, ed. Helmut Norpoth, Michael Lewis-Beck, and Jean-Dominique
Lafay. Ann Arbor: University of Michigan Press, 123–
39.
Lebo, Matthew J., and Daniel Cassino. 2007. “The Aggregated
Consequences of Motivated Reasoning and the Dynamics
of Partisan Presidential Approval.” Political Psychology 28:
719–46.
Lewis-Beck, Michael, and Mary Stegmaier. 2008. “The Economic Vote in Transitional Democracies.” Journal of Elections, Public Opinion, & Parties 18: 303–23.
MacKuen, Michael, B., Robert S. Erikson, and James A. Stimson. 1992. “Peasants or Bankers? The American Electorate
and the US Economy.” American Political Science Review 86:
597–611.
Mansfield, Edward D., and Jack Snyder. 1995. “Democratization and the Danger of War.” International Security 20: 5–
38.
Mietzner, Marcus. 2009. “Indonesia in 2008: Yudhoyono’s
Struggle for Reelection.” Asian Survey 49: 146–55.
Miller, Arthur H., William Reisinger, and Vicki L. Hesli. 1996.
“Understanding Political Change in Post-Soviet Societies: A
Further Commentary on Finifter and Mickiewicz.” American Political Science Review 90: 153–66.
Mishler, William, and John Willerton. 2003. “The Dynamics
of Presidential Popularity in Post-Communist Russia: Cultural Imperative versus Neo-Institutional Choice?” Journal
of Politics 65: 111–41.
Morin, Richard, and Nilanthi Samaranayake. 2006. The Putin
Popularity Score: Increasingly Reviled in the West, Russia’s
Leader Enjoys Broad Support at Home. Washington, DC: Pew
Research Center.
Mueller, John. 1973. War, Presidents, and Public Opinion. New
York: Wiley.
Remnick, David. 2007. “Letter from Moscow: The Tsar’s Opponent.” The New Yorker, October 1, 2007.
609
Robinson, Peter M. 1995. “Gaussian Semiparametric Estimation of Long Range Dependence.” Annals of Statistics 23:
1630–61.
Rose, Richard. 2007a. “Going Public with Private Opinions: Are
Post-Communist Citizens Afraid to Say What They Think?”
Journal of Elections, Public Opinion, and Parties 17: 123–42.
Rose, Richard. 2007b. “The Impact of President Putin on Popular Support for Russia’s Regime.” Post-Soviet Affairs 23:
97–117.
Rose, Richard, William Mishler, and Neil Munro. 2004. “Resigned Acceptance of an Incomplete Democracy: Russia’s
Political Equilibrium.” Post-Soviet Affairs 20: 195–218.
Rzyhkov, Vladimir. 2004. “The Liberal Debacle.” Journal of
Democracy 15: 52–58.
Sanders, David. 2000. “The Real Economy and the Perceived
Economy in Popularity Functions: How Much Do Voters
Need to Know? A Study of British Data, 1974–97.” Electoral
Studies 19: 275–94.
Sil, Rudra, and Cheng Chen. 2004. “State Legitimacy and the
(In)significance of Democracy in Post-Communist Russia.”
Europe-Asia Studies 56: 347–68.
Tortorice, Daniel L. 2009. “Unemployment Expectations and
the Business Cycle.” Brandeis University. Unpublished
manuscript.
Tucker, Joshua A. 2006. Regional Economic Voting: Russia, Poland, Hungary, Slovakia and the Czech Republic,
1990–1999. New York: Cambridge University Press.
Weyland, Kurt. 2003. “Economic Voting Reconsidered: Crisis
and Charisma in the Election of Hugo Chávez.” Comparative
Political Studies 36 (September): 822–48.
White, Stephen, and Ian Mcallister. 2008. “The Putin Phenomenon.” Journal of Communist Studies and Transition
Politics 24: 604–28.
White, Stephen, Sarah Oates, and Ian McAllister. 2005. “Media Effects and Russian Elections, 1999.” British Journal of
Political Science 35: 191–208.
Whitefield, Stephen. 2005. “Putin’s Popularity and Its Implications.” In Leading Russia: Putin in Perspective, Essays in
Honour of Archie Brown, ed. Alex Pravda. New York: Oxford
University Press, 139–60.
Wyman, Matthew. 1997. Public Opinion in Postcommunist Russia. New York: St. Martin’s Press.
Yeltsin, Boris. 2000. Midnight Diaries. New York: Public Affairs.